Download Document

ESSENTIALS OF ASHÉNINKA PERENÉ GRAMMAR
by
Elena Mihas
Este arquivo, fornecido pela autora em dezembro de 2010,
está disponível para download no seguinte endereço:
http://www.etnolinguistica.org/tese:mihas_2010
A Dissertation Submitted in
Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy
In Linguistics
at
The University of Wisconsin-Milwaukee
December 2010
ESSENTIALS OF ASHÉNINKA PERENÉ GRAMMAR
by
Elena Mihas
A Dissertation Submitted in
Partial Fulfillment of the
Requirements for the Degree of
Doctor of Philosophy
In Linguistics
at
The University of Wisconsin-Milwaukee
December 2010
________________________________________________________________________
Major Professor
Date
________________________________________________________________________
Graduate School Approval
Date
ii
ABSTRACT
ESSENTIALS OF ASHÉNINKA PERENÉ GRAMMAR
by
Elena Mihas
The University of Wisconsin-Milwaukee, 2010
Under the Supervision of Fred Eckman
The objective of this dissertation is to present a preliminary grammatical account of
Ashéninka Perené, an endangered Arawak language of Southeastern Peru. The
description and analysis of the language is based on the 29-week field research
conducted in an area of the Southwest Amazonian high jungle. Interesting issues of
Ashéninka Perené grammar include a lack of phonemic voice distinction in stops,
fricatives, and affricates, a series of palatalized stops, and a set of palatalized alveolar
phonemes. Grammatical morphemes exhibit a great deal of multifunctionality. Possession
is marked on the possessee or by juxtaposition. Non-masculine is the default gender.
There is a small underived adjective class with mere thirteen members. The nominal
classification system is fairly extensive. There is an array of highly productive applicative
derivations. Aspectual and modal systems are very complex. Ashéninka Perené shows a
straightforward correspondence between unreal events and their encoding in the language
as irrealis. Negated irrealis is expressed on the verb as realis. There is an abundance of
clausal conjunctions. The language has a nominative-accusative system of grammatical
iii
alignment combined with the frequently occurring split intransitive phenomenon which is
grammatically and pragmatically conditioned.
_______________________________________________________________________
Major Professor
Date
iv
© Copyright by Elena Mihas, 2010
All rights reserved
v
TABLE OF CONTENTS
1. The language and its speakers
1.1 Linguistic profile of Ashéninka Perené
1.2 Ashéninka Perené within the Kampan subgroup and the Arawak
language family
1.3 The regional history and status of Ashéninka Perené
1.4 Prior research on the language
1.5 Fieldwork materials, research methods, and consultants
2. Phonology
2.1 Vowels
2.1.1 Short vowels
2.1.2 Long vowels and diphthongs
2.2 Consonants
2.2.1 Phoneme inventory
2.2.2 Stops /p/, /t/, /k/ and their palatalized counterparts /pʲ/, /tʲ/,
and /kʲ/
2.2.2.1 The phonemes /p/ and /pʲ/
2.2.2.2 The phonemes /t/ and /tʲ/
2.2.2.3 The phonemes /k/ and /kʲ/
2.2.3 Affricates ts /tsʰ/, tz /ts/, and ch /ʧ/
2.2.4 Alternative analyses of palatalized segments and affricates
2.2.5 Fricatives s/s/, sh /ʃ/, and h /h/
2.2.6 Nasals n /n/, m /m/, ny /ɲ/, and the archiphoneme N
2.2.7 Liquids r/ɾ/ and /ɾʲ/
2.2.8 Semivowels v /w/ and y /j/
2.3 Sounds in Spanish loanwords
2.4 Phonotactics
2.4.1 Syllable structure
2.4.2 The minimal word
2.1.3.3 Stress
2.5 Intonation
2.6 Phonological differences among Ashéninka Perené
speech communities
2.7 Orthography
2.8 Morphophonemics
2.8.1 Lenition
2.8.2 Epenthesis
2.8.3 Deletion
2.8.4 Sibilant neutralization
3. Morphology
3.1 Grammatical word vs. phonological word in Ashéninka Perené
3.2 Morphological typology
3.3 Formative classes
vi
1
4
14
16
26
28
44
44
45
50
54
54
55
55
56
57
58
59
60
62
63
64
68
70
72
73
75
77
79
81
85
86
87
88
89
91
91
95
97
3.3.1 Definitions of free and bound morphemes, roots,
and formatives
3.3.2 Core formatives: affixes
3.3.3 Peripheral formatives: clitics and particles
3.3.4 Other morphological processes
3.4 Word classes
3.4.1 Nouns
3.4.2 Verbs
3.4.3 Adjectives
3.4.4 Adverbs and adverbial expressions
3.4.5 Pronouns
3.4.6 Numerals
3.4.7 Interjections, ideophones, and discourse particles
4. Clause linking
4.1 Relative clauses
4.1.1 The =ri/=ni relativizing strategy in headed relative clauses
4.1.2 Relativzing strategies in headless clauses
4.1.3 Other relativizing strategies
4.1.4 Other functions of the forms =ri and =ni
4.2 Complementation
4.2.1 The grammatical profile of the complement clause
4.2.2 Semantic types of complement-taking verbs
4.2.2.1 Complement-taking primary verbs
4.2.2.2 Complement-taking secondary verbs
4.3 Adverbial clauses
4.3.1 Temporal clause linking
4.3.2 Conditional clause linking
4.3.3 Consequence clause linking
4.4 Other types of clause linking
97
102
108
115
121
121
126
142
151
159
184
185
191
191
193
201
204
210
217
220
225
227
235
244
245
257
265
280
References & Bibliography
Appendix A. Basic word list
Appendix B. Text samples
B.1 Narrative
B.2 Song
B.3 Riddle
B.4 Conversation
Curriculum Vita
286
296
299
299
312
314
314
320
vii
LIST OF FIGURES
Maps:
Map 1
Map 2
Figures:
Figure 1
Figure 2
Ashéninka Perené settlements
Kampan languages
2
3
Figure 8
Figure 9
Figure 10
Figure 11
Figure 12
Figure 13
The Pampa Michi historian Raul Bernata
19
Transcript of a story about a clash of Upper Perené Indians
with colonists
19
A hill slope with chacras in the Perené District
22
A craft stand in Pampa Michi
22
Language consultants from Bajo Marankiari and Pampa Michi
43
The devoiced high front vowel [i̥ ] in kapicheeni ‘little’
46
Medial and word-final realization of the phoneme [e]
in shetene ‘afternoon’
46
Formant plots of vowels of female and male speakers from PM
48
Formant plots of vowels of female and male speakers from BM
48
Articulation of the bilabial approximant in ivito ‘his canoe’
62
Phonetic realization of the bilabial approximant in pava ‘god’
65
Phonetic realization of the bilabial approximant in novori ‘my leg’ 65
Phonetic realization of the bilabial approximant in ivito ‘his canoe’ 66
Figure 14
Figure 15
The palatalized stop /tʲ/ in novityero ‘I will fell it’
Continuum of headedness of RC
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
viii
68
204
LIST OF TABLES
Table 1
Table 2
Table 3
Table 4
Table 5
Table 6
Table 7
Table 8
Table 9
Table 10
Table 11
Table 12
Table 13
Table 14
Table 15
Table 16
Table 17
Table 18
Table 19
Table 20
Table 21
Table 22
Table 23
Table 24
Table 25
Table 26
Table 27
Table 28
Table 29
Table 30
Table 31
Table 32
Table 33
Table 34
Table 35
Table 36
Table 37
Table 38
Table 39
Table 40
Table 41
Table 42
Vowel phonemes
Phonetic realizations of short vowels
Diphthongs
Consonant phonemes and their allophones
Ashéninka writing systems developed from the 1980s to date
Verbal suffixes
Nominal suffixes (an overview)
Class terms
Nominal classifiers
Suffixes which occur with multiple word classes
Verbal proclitics
Verbal enclitics
Plant-based parts in compounds
Gender agreement patterns
Summary of the cumulating reality status/reflexivity
values of verbs
Lexical classes of verbs
Existential/possessive verbs
Copulas
A set of underived adjectives
Properties of underived adjectives
Properties of derived adjectives
Temporal adverbs
Locative-existential adverbs
Simple personal pronouns
Four sets of personal pronouns
Possessive pronouns
Forms and functions of bound pronominal forms
Allomorphy of bound pronominal forms
Nominal demonstrative pronouns
Indefinite pronouns
Elicited numeral forms
Interjections
Ideophones
Discourse particles
Summary of clause linkage
Syntactic functions of the common argument in MC and RC
Types of relative clauses
Synchronic uses of the ri and ni forms
Primary complement-taking verbs
Secondary complement-taking verbs
Temporal clause linking devices
Possible and counterfactual clause linking devices
ix
45
45
51
55
83
104
106
106
106
108
111
112
118
122
131
133
137
139
143
146
150
153
156
160
161
168
169
170
174
182
185
186
188
189
192
197
209
217
228
236
246
258
Table 43
Table 44
Consequence clause linking devices
Contrast, disjunction, manner linking devices
x
266
280
LIST OF ABBREVIATIONS
A
ABS
ADJ
ADV
ADV.P
APPL
APPL.BEN
APPL.INST
APP.INT
APPL.PRES
APPL.REAS
APPR
AUG
CAUS
CAUS.AGT
CAUS.SOC
CC
CL
CNT.F
COM
CONN
COND
COP.C
CUST
DE
DEM
DIM
DIR
DIST
DUB
DUR
EMPH
EP
EXCL
EXIST
FOC
FOC.O
FRUS
HAB
ICPL
IDEO
IMP.P
INCH
subject of transitive clause
absolutive
adjectivizer
adverbial
adverbial past
generalized applicative
applicative benefactive
applicative instrumental
applicative of intent
presential applicative
applicative of reason
apprehensive
augmentative
causative
causative agentive
causative sociative
coordinate clause
classifier
counterfactual
comitative
connective
conditional
copula of capacity
customary aspect
dependent event
demonstrative
diminutive
directional
distributive
dubitative
durative
emphatic
epenthetic
exclamative (general focus)
existential
focus
focused object
frustrative
habitual
incompletive
ideophone
impersonal passive
inchoative
xi
INST
IRR
IRR.REFL
LOC
m.
MC
M.CAUS
N.C.FOC
NEG
NEG.EXIST
NEG.IRR
NEG.P
NEG.REAL
n.m
NMZ
NP
O
OPT
PL
poss
PP
PRF
PROG
Q
RC
RCP
REGR
REL
REP
REV
S
SG
SEP
STAT
STAT.T
VP
WH
1
2
3
<>
~~
instrumental
irrealis
irrealis reflexive
locative
masculine
main clause
malefactive causative
non-contradictory focus
negative
negative existential
negative irrealis
negative polarity
negative realis
non-masculine
nominalizer
noun phrase
object of transitive clause
optative
plural
possessive
positive polarity
perfective
progressive
interrogative
relative clause
reciprocal
regressive
relativizer
repetitive
reversative
subject of intransitive clause
singular
separative applicative
stative
temporal stativity
verb phrase
content interrogative
first person
second person
third person
infix
reduplicated element
xii
ACKNOWLEDGMENTS
My profound gratitude is extended to my Ashéninka Perené teachers Gregorio Santos
Pérez (Villa Perené), Delia Rosas Rodríguez (Bajo Marankiari), Raul Martin Bernata
(Pampa Michi), and Bertha Rodríguez de Caleb (Bajo Marankiari) whose comprehensive
knowledge of the language enormously contributed to my understanding of the
language’s grammar and Ashéninka culture. I also want to thank other speakers with
whom I worked closely: Ines Pérez de Santos (Bajo Marankiari), Moises Santos Rojas
(Bajo Marankiari), Abdias Caleb Quinchori (Bajo Marankari), Victoria Manchi de Martin
(Pampa Michi), Ernesto Manchi Lopez (Pampa Michi), Daniel Bernales Quillatupa
(Aroshi) (La Merced), Paulina Caleb de Leon (Bajo Marankiari), Paulina García Ñate
(Bajo Marankiari), Alberto Pérez Espinoza (Bajo Marankiari), and Luis Manchi Lopez
(Santari). Special thanks go to the chief of Bajo Marankiari Osbaldo Rosas Rodríguez.
His support of the language documentation and grammar-writing project has allowed me
to fully engage native speakers of Bajo Marankiari which served as the base research
community.
I also want to acknowledge the enlightening influence of the late Professor of
Linguistics Michael Noonan who trained me as a linguist and advised me as my mentor
during 2004-2009. I mainly owe to him whatever I was able to accomplish as a
documentary linguist. Other scholars for whom I want to express my deepest appreciation
are Edith Moravcsik and Lev Michael. Their insightful comments on my work have
immensely benefited this doctoral project. I am particularly grateful to Lev Michael who
suggested that I do research on an Ashéninka variety. My sincere acknowledgements are
extended to the Arawakanists whom I have never met but whose excellent work built a
xiii
foundation for my research and inspired me to continue in their footsteps: Judith Payne,
David Payne, Ronald Anderson, Mary Ruth Wise, and Alexandra Aikhenvald. My
analysis of data largely drew on the invaluable guidelines of the Basic Linguistic Theory
by R.W.M. Dixon, whom I would like to acknowledge with gratitude. I owe a great deal
of thanks to the Chairman of my doctoral committee Fred Eckman for creating a
supportive climate at the Linguistics Department at UW-Milwaukee which helped me
survive the difficulties of conducting field research in a foreign country. My thanks go to
Christopher Roth for his input on the final drafts of the grammar. During the entire
project, the equipment needs and technical issues were attended to by Jim Cavanaugh and
Jeff Loomis, for which I am grateful too.
I am very grateful to the organizations that funded my research: the National Science
Foundation (Grant #0901196) and the Hans Rousing Endangered Languages Project
(HRELP) (Grant SG0002).
Warmest thanks go to my family in Russia, my friends in Peru and the US who have
helped me in many ways during 2009-2010: Aldo Sedano Marro, Vicki and Tad Johnson,
Pat Costello, Sally Noonan, Mike Darnell, Amy Callahan, and Elena Rucker. Many
thanks to the sisters Carmen and Lorena Broq Tremolada from fundo San Jose in La
Merced who kindly adopted me during the fieldwork periods in 2009-10.
Above all, I want to thank my dear husband Pete Mihas whose faith in me and
unwavering dedication made it possible for me to complete the project.
xiv
1
Chapter 1
The language and its speakers
Ashéninka Perené speakers live in the dispersed settlements along the Upper Perené
River in Distritos of Chanchamayo and Perené of Departamento Junín, Peru. The speaker
base is reported to be at c.5,500 (Gordon 2005) but these data are questionable. The
problem is the misinterpretation of population size as the number of speakers. Fieldwork
results have revealed that the actual number of speakers is significantly lower, c. 1,000.
The number of Ashéninka Perené-speaking communities is estimated at 37 by the local
government of Distrito Perené (http:// www.muniperene.gob.pe). However, fieldwork
data show that the total number of settlements is 35: Pampa Michi, San Miguel Centro
Marankiari, Bajo Marankiari, Villa Perené, Bajo Esperanza, Alto Esperanza, Mariscal
Cáceres, San Jerónimo de Urinaki, Pucharini, Chirani, Shinari, Cumbre San Roman,
Satinaki, Alto Incariado, Bajo Incariado, Ichatingari, Santa Rosa Ubiriki, Churingaveni,
Pachacutec, Campo Verde, Alto Pucharini, Centro Pumpuriani, Kivinaki, Alto Kivinaki,
Cerro Picaflor, Shankivironi, Shivitsari, Shintoriato, Bajo Aldea, Sotani, Huacamayo, San
Cristobal, Alto Huacamayo, San Jose Kuviriani, San Martin de Ubiriki. They are shown
in Map 1. Approximately half of the villages are located on the valley floor; the rest are
nestled in the hills on both sides of the river.
2
MAP 1. Ashéninka Perené settlements in the Districts of Chanchamayo and Perené
The socio-linguistic data (numbers of speakers and settlements) have been
contributed by the following native speakers of Ashéninka Perené: Gregorio Santos Pérez
(Villa Perené), Elias Meza Pedro (Mariscal Cáceres), Carmen Santos Lopez (Alto
Esperanza), Marino Samaniego Domingo (San Miguel Centro Marankiari), Osbaldo
Rosas Rodriguez (Bajo Marankiari), Cristobal Jumanga (Pucharini), Enrique Martinez
(Bajo Chirani), Mario Prado Lopez (Santa Rosa de Uviriki), Delfin Pongo (Alto
Incariado), Adan Santo Castañeda (Cumbre San Roman), Jorge Quinchori (San Jerónimo
de Urinaki), Lucas Florez (Renacimiento Kivinaki), Ricardo Kamacho (Pampa Michi),
Nestor Ramirez (Santo Domingo), Nilda Camiñiri (Shinari), Elmer Quinchori (Aldea
Bajo), Victor Camacho ( Shintoriato), Adolfo Gutierrez (Churigaveni), Dany Robles
3
(Huacamayo), Adolfo Marcos (Alto San Luis), Dionicio Minkori (Bajo Esperanza),
Elmer Antonio (Camonashari), Mateo Cristobal (Platanillo Shimaki), Daniel Bernales
Quillatupa (La Merced).
The river valley, called Alto Perené (Upper Perené River), is flanked by the
settlements of Ashéninka Pichis in the north, Ashéninka Gran Pajonal in the north east,
and Asháninka Tambo-Ene in the east. Map 2 uses the following numerical symbols for
the areas inhabited by speakers of Kampan languages: 3-Ashéninka Apurucayali, 9Ashaninka Tambo-Ene, 10-Ashéninka Pajonal, 11-Ashéninka Perené, 19-Kakinte, 46Machiguenga, 53-Nanti, 55-Nomatsiguenga, 65-Ashéninka Pichis, 80-Ucayali-Yurúa.
MAP 2. Kampan languages (an abridged SIL 2004 Map of Southern Peru, from www.ethnologue.com)
The language described in this grammar is that spoken in the villages of Pampa Michi
(Chanchamayo District) and Bajo Marankiari (Perené District).
4
1.1 Linguistic profile of Ashéninka Perené
Asheninka Perené is a polysynthetic, head-marking, agglutinating, mainly suffixing
language with rich verbal morphology. It has eighteen consonants and eight vowels.
There is a set of palatilzed stops and a series of palatalized alveolar consonants for all
manner of articulation (a stop, affricates, a fricative, a liquid, and a nasal) which contrast
with their non-palatalized counterparts. The language lacks phonemic voice distinction;
its stops, fricatives and affricates are voiceless. Neither does it have aspirated consonants.
There are two sibilant fricatives s and ʃ; one glottal fricative h; two liquids with a flap
articulation ɾ and its palatalized counterpart ɾʲ; two glides, the bilabial approximant w and
the palatal glide j. It has three nasal stops m, n, and ɲ which contrast with the
underspecified nasal N. The vowels i, e, a, o have long counterparts. The canonical
syllable structure is CV. The minimal word must have two syllables and an independent
stress. Phonological processes (lenition, sibilant neutralization, epenthesis, vowel
deletion) occur on the boundaries between a prefix and a root, a root and a suffix, a root
and a root, and a suffix and a suffix.
Open word classes are verbs, nouns, and derived adjectives. Underived adjectives
form a small closed class of circa thirteen members. Adverbs form a semi-closed class
(e.g. place and locative-existential adverbs are closed subclasses while it is possible to
derive some time and modal adverbs from verbs). Demonstratives, possessives, indefinite
forms, personal pronouns, interrogatives are closed classes. Members of open classes can
function as predicates; however, only verbs exhibit unrestricted morphological
possibilities (e.g. nouns, adjectives, and adverbs do not occur in the imperative mood).
5
Structural properties of nouns involve the categories of case (locative case -ki),
morphologically expressed gender (-ri ‘masculine, -ro ‘non-masculine’), optionally
expressed plural number –paye, possession, locality (=ka, =ra, =Nta), and a crosslinguistically infrequent nominal past tense =raNki. Derivational categories include class
terms, classifiers, and degree (diminutive –aniki, -patsaini, -peta, and augmentative
–santsa, -mashi). There are at least twelve class terms, mostly plant-based, e.g. –chee
‘thorn’, -paNki ‘plank’, etc. and ten classifiers. Most of the classifiers occur only with
nouns while a few have been attested with verbs. Possession is morphologically
expressed on two groups of nouns: obligatorily (inalienably) possessed and optionally
(alienably) possessed. There are also non-possessable nouns which always appear bare
without any inflectional morphology. When an NP is marked for possession, only the
possessed NP takes possessive inflection, e.g. kaminkari ichenkoporoki kamiNka-ri
i=cheNko-poroki [dead-NZR 3m.poss=pants-CL: quantity] ‘the dead person’s pants’.
Obligatorily possessed nouns take person/gender clitics (gender distinction is only made
when the possessor is the 3SG entity) while optionally possessed nouns take
person/gender clitics and a possessive suffix –ni, -te, or –ri, e.g. ishintsipaari
i=shintsipaa-ri [3m.poss=raft-poss] ‘his raft’. Possession is also expressed by
juxtaposition, which occurs when the possessor is a deverbal nominalization, e.g.
ashitarori eentsipaye ashit-a=ro=ri eentsi-paye [have-REAL= 3n.m.O=REL child-PL]
‘owners of children [parents]’, the possessor NP appears before the possessed NP. When
optionally-possessed nouns appear in an ‘unpossessed’ form, they take an unspecified
possessor suffix –(n)tsi or -rintsi/-rontsi. There is no case marking on core constituents
(A, S, O); the only case-marking locative suffix –ki can have any of the following spatial
6
meanings: ‘in’, ‘at’, ‘on’, ‘by’, ‘to’. The diffuse spatial meaning of the suffix -ki can be
extended to a temporal sense ‘during/over’.
Verb roots are always bound and minimally have a subject proclitic and a reality
status suffix. They typically appear with other affixes which occupy multiple structural
slots for number, classifiers, valence-adjusting/preserving, adverbial, and
aspectual/reality status categories. In addition, there are slots filled with personal,
subordination (temporal/spatial =ra ‘when/where’, relative =ri/=ni), mood categories
(interrogative =ka, dubitative =ma, conditional =rika, counterfactual =mi, optional =ta,
apprehensive =kari), adverbial past tense =raNki, adverbial locative =ka, =ra, =Nta, and
pragmatic clitics =ha, =tya, =kya, =ve. The language makes a formal distinction between
real and unreal events. The scope of the irrealis morphemes -e and –ia covers the entire
notional range of what is defined as irrealis (unrealized) events: future events (1.1),
imperative (1.2), optative (1.3), counterfactual (1.4), conditional (1.5), purposive (1.6),
some types of complements (1.7), negated realis clauses (1.8), deontic modality
constructions (1.9).
FUTURE
1.1
Nosaitatyero nihaa.
no=sai-t-aty-e=ro
nihaa
1SG.A=pour-EP-PROG-IRR=3n.m.O water
‘I will pour the water.’
1.2
Pamine mapi.
p=amin-e
mapi
2S=look.for-IRR stone
‘Look for a stone.’
IMPERATIVE
7
OPTATIVE
1.3
Apaata novetsikayetavaketa omorona.
apaata no=vetsik-a-ye-t-av-ak-e=ta
o=morona
later 1SG.A=make-EP-DIST-EP-DIR-PRF-IRR=OPT 3n.m.poss=hole
‘Later I will make holes [for the seedlings].’
COUNTERFACTUAL
1.4
Airorika itsonkahetakaemi.
airorika i=tsoNk-a-he-t-ak-a-e=mi
or
3m.A=kill-EP-PL-EP-PRF-REGR-IRR=CNT.F
‘Or he would have killed them [but he didn’t].’
CONDITIONAL
1.5
Arika inyakero apankite kaniri, yamakeri ivaakate.
arika
i=ny-ak-e=ro
a=paNki-t-e
kaniri
when/if 3m.A=see-PRF-IRR=3n.m.O 1PL.S=plant-EP-IRR manioc
y=am-ak-e=ri
i=vaaka-te
3m.A=bring-PRF-REAL=3m.O 3m.poss=cow-poss
‘When/if they see us plant manioc, they will bring their cows.’
PURPOSIVE
1.6
Ikeni ivametairi onkantya yamitakoteri.
i=ken-i
i=vamet-a-e=ri
oNkaNtya
3m.S=come-REAL 3m.A=teach-REGR-IRR=3m.O in.order
y=amitako-t-e=ri
3m.A=help-EP-IRR=3m.O
‘He came to teach them [young men] so that they could help him [Juan Santos
Atahualpa].’
COMPLEMENT CLAUSE
1.7
Nokovatzi nopankite mapocha.
no=kov-atz-i
no=paNki-t-e
mapocha
1SG.S=want-PROG-REAL 1SG.S=plant-EP-IRR papaya
‘I want to plant papaya.’
NEGATED REALIS CLAUSE
1.8
Te nonimopirotya noshinkipirotya.
te
no=nimo-piro-t-ia
no=shiNki-piro-t-ia
NEG.REAL 1SG.S=like-AUG-EP-IRR 1SG.S=get.drunk-AUG-EP-IRR
‘I don’t like to drink alcohol.’
8
DEONTIC MODALITY CONSTRUCTION
1.9
Ontzimatye antavaite maaroiteni.
ontzimatye
aNt-a-vai-t-e
be.necessary work-EP-DUR-EP-IRR
‘We should work all together.’
maaro<ite>ni
all<AUG>ADV
There are two lexical classes of verbs, class A and class I, labeled on the basis of the
reality status inflectional endings they take. Membership in each class is largely
unpredictable, except for reflexive verbs which all belong to class A. In addition to
regular verbs, there are two sets of morphologically defective verbs: existentials/
possessives and copulas. The existential/possessive verbs include positive polarity verbs
tzimatsi ‘existential/possessive’, ainiro ‘existential/possessive’, kant ‘be’, and the
negative polarity verb tekatsi ‘negative existential/possessive’. The set of copulas is
comprised of the existential copula na, copula of naming pai, copula of location saik,
copula of capacity kara, and the negative copula kaari. Verbs divide into two basic
transitivity classes depending on the transitivity type of clauses they occur in: intransitive
clauses with one obligatory argument S (intransitive subject) and transitive clauses which
have two arguments A (transitive subject) and O (transitive object). Ditransitive clauses
are infrequent. Verb roots are essentially ambitransitive; they can have either transitivity
value. For example, the verb roots kaa ‘bathe’and kotsi ‘cook’ may appear in ditransitive
clauses, as seen in (1.10). The number of participants on the scene is three: the mother
who is assisting her daughter during and after labor, the daughter, and the newborn baby.
The objective form =ni, which is a third person singular enclitic (unspecified for gender),
appears in the benefactive role. The transitive subject A is cross-referenced on the verb
9
by the third person non-masculine proclitic o=, the transitive object O is marked by the
third person masculine enclitic =ri.
1.10
Onkaatakeniri, onkotsiteniro oyari.
o=N-kaa-t-ak-e=ni=ri
3n.m.A=IRR-bathe-EP-PRF-IRR=3.O=3m.O
o=N-kotsi-t-e=ni=ro
oyari
3n.m.A=IRR-cook-EP-IRR=3O=3n.m.O her.food
‘She [the mother] will bathe him [the baby] for her [for the daughter], she will
cook food for her.’
Ashéninka Perené does not have a valence-decreasing passive derivation although
there is an impersonal passive marker –ai. Noun-incorporation, reciprocal –av and
reflexive –a~-eya are valence-decreasing operations; valence-increasing mechanisms
include applicative derivations (benefactive –veNt, instrumental/reason –aNt, separative
–pitsa, presential –imo, generalized applicative –ako, applicative of intent –ashi),
customary –aNt, and causativization. The three causative morphemes include the
malefactive causative prefix mi-, the non-productive agentive causative prefix o-~oi-~ov-,
and the productive causative sociative suffix –ak. A periphrastic causative construction,
which has a connotation of moral obligation, involves the verb root shintsi ‘be strong’,
the benefactive applicative –veNt~-viNt, as seen in (1.11); cf. the causative suffix -ak in
(1.12).
1.11
Ishintsivintakina nayeri shiramparika.
i=shintsi-viNt-ak-i=na
n=ay-e=ri
shiraNpari=ka
3m.A=be.strong-BEN-PRF-REAL=1SG.O 1SG.A=take-IRR=3m.O man=DEM
‘They forced me to marry this man.’
10
1.12
Ivakakinari iyoka.
i=v-ak-ak-i=na=ri
iyoka
3m.A=kill-CAUS-PRF-REAL=1SG.O=3m.O DEM
‘They made me kill him.’
The language makes aspectual distinctions (perfective –ak, regressive –ah,
incompletive –it, inchoative –ima(in), progressive –aty~-atz, durative –vai, habitual
–apiNt, customary –aNt, repetitive –a (‘continuous repetition’) and iterative –apanaNt
(‘one more time’), reality status (realis –a, -i; irrealis –ia, -e), and mood distinctions
(interrogative =ka, dubitative =ma, conditional =rika, counterfactual=mi, optional =ta,
apprehensive =kari). In addition to two aspectual adverbs tekira ‘not yet’ and aikiro
‘still’, there is a set of adverbial verbal suffixes describing manner (-ite ‘quickly’,
‘surprisingly’), time (–it ‘before’, -aman ‘early’), and degree (–niNt ‘a little bit’,
‘slightly’, -pero ‘indeed’, ‘truly’) of an action. There is also an optional adverbial past
tense clitic =raNki which denotes that the action took place at some point in the past, as
seen in (1.13).
1.13
Nohatanakiranki.
no=ha-t-an-ak-i=raNki
1SG.S=go-EP-DIR-PRF-REAL=ADV.P
‘I went [at some point in the past].’
Canonical (i.e. second person) imperatives, expressing commands, are formally
undistinguishable from polite requests or hortatives since the verb obligatorily takes
subject person marker and irrealis inflection to express an unrealized action, as seen in
(1.13). The force of command is intensified by the exclamative clitic =ve. Other types of
imperatives are first person cohortatives (‘let’s do it’), second person hortative (‘you’ll do
11
that, right?’), second and third person exhortative (‘beware of doing it’) and second
person entreaties or pleas (‘please please do this’). Examples are provided in (1.13-1.17).
COMMAND
1.13
Pikantenaro.
pi=kaNt-e=na=ro
2A=say-IRR=1SG.O=3n.m.O
‘Say it to me.’
COHORTATIVE
1.14
Tsame ahate!
tsame
a=ha-t-e
come.on 1PL.S=go-EP-IRR
‘Let’s go.’
HORTATIVE
1.15
Pimpokera.
pi=N-pok-e=ra
2S=IRR-come-IRR=ADV
‘You will come, right?’
EXHORTATIVE
1.16
Yayimikari!
y=ay-i=mi=kari
3m.A=take-REAL=2.O=APPR
‘He shouldn’t take [it] from you!’
ENTREATY
1.17
Pintyankahaitenami!
pi=N-tyaNk-ah-a-it-e=na=mi
2A=IRR-send-REGR-EP-ICPL-IRR=1SG.O=CNT.F
‘Please send me [the money]!’
There is one negative imperative construction formed with the irrealis negator
airo/ero. The negated imperative is formally equivalent to the negated irrealis clause. To
indicate the force of the prohibitive, the reduplicated ending –tzi is used, as seen in
(1.18).
12
PROHIBITIVE
1.18
Ero paminavaitzitzi!
airo
p=amin-a-vai~tzitzi
NEG.IRR 2S=look-EP-DUR~EP.REAL
‘Don’t look!!’
Ashéninka Perené exhibits periphrastic negation. The set of negators includes the
negative realis particle te, negative irrealis particle airo~ero, negative polarity verb
tekatsi ‘existential/possessive’, the copula kaari ‘negative existential’, and the aspectual
adverb tekira ‘not yet’. Their position is fixed; the negators always precede the predicate.
The standard negation strategy, which is a basic means of expressing a negated
declarative clause, involves the negators te and airo~ero. The particle te is used to negate
realis clauses whereas airo~ero negates irrealis propositions. Both negators have scope
over the reality status of the negated clause: te-negated clauses obligatorily take irrealis
marking and airo-negated clauses take realis marking.
The language has various relativizing strategies attested with canonical and noncanonical relative clause constructions. The canonical relative clause construction
involves marking of the predicate of the headed relative clause by the relativizers =ri or
=ni; =ri relativizes the common argument in A, S, O function in realis clauses and in O
function in irrealis clauses while =ni is associated with relativization of the common
argument in A and S function in irrealis clauses. Ashéninka Perené lacks a complement
clause construction but complement clauses can be treated as such based on the
functional definition of complementation. The semantic types of adverbial subordinate
clauses include temporal, conditional, purposive, causative, resultative, motion purposive,
undesirable consequence. Two other types - counterfactual and manner adverbial clauses
- are coordinated. This extensive set of adverbial subordinate clauses is marked by
13
construction-specific free and/or bound grammatical morphemes (arika ‘when/ if’, =ra
‘when’, ‘where’, okaNta/ikaNta ‘in the meantime’, tekira ‘before’, irohatzi ‘until’,
aririka ‘if’, =rika ‘when/if’ , tema, kama ‘because’, irotaki/iritaki + -aNt ‘that’s why’, aNt + =ri ‘in order’, oNkaNtya/okaNta ‘so that’, =kari ‘lest’).
The language exhibits nominative-accusative grammatical alignment with traces of
split intransitivity when intransitive subject is realized by O person markers and occupies
the verb post-stem person enclitic slot. It appears that this non-canonical realization of
intransitive subjects is speaker-dependent, pragmatically motivated, and grammatically
conditioned, typically by the aspectual formatives perfective –ak or regressive -ah which
have a terminative connotation. Topic continuity1 is highlighted by marking intransitive
subjects as objects, as seen in (1.19-1.20).
1.19
Hatahana.
ha-t-ah-a=na
go-EP-REGR-REAL=1SG.O
‘I am going (so long).’
1.20
Iro kamatsavaitakina.
iro kamatsa-vai-t-ak-i=na
but crawl-DUR-EP-PRF-REAL=1SG.O
‘But I continued to crawl on the floor.’
The basic word order is VAO (verb-transitive subject-object) but AVO (transitive
subject-verb-object) is also possible. The order of the fully expressed verb arguments
largely depends on their pragmatic status, with topical elements appearing in the
preverbal position. The focused constituents can either precede or follow the verb.
1
Payne & Payne observe that O-perfective verbs (i.e. verbs which bear the perfective suffix -ak and whose
subjects are marked as an object) indicate that the subject participants are topically continuous referents;
these verbs tend to mark the storyline in narratives and cluster at the climax (2005:48-50).
14
A number of aspects of Ashéninka Perené grammar are cross-linguistically infrequent
such as exuberant applicative derivations, spectacular polyfunctionality of its
grammatical morphemes, extremely complex aspectual and modal systems, the past tense
nominal morpheme, pragmatically conditioned split intransitivity, expression of irrealis
negation as realis, and the straightforward correspondence between conceptually unreal
events and their encoding in the language as irrealis.
1.2 Asheninka Perené within the Kampan subgroup and the Arawak2 language
family
Ashéninka Perené belongs to the Kampan subgrouping of South Arawak, along with
other varieties of Ashéninka, Ashaninka, Caquinte (Kakinte), Nomatsiguenga,
Machiguenga, and Nanti.
The number of Ashéninka varieties listed is somewhat arbitrary in that it is difficult to
say whether some varieties should be considered dialects or languages. Linguistic
scholarship distinguishes several dialects of Ashéninka (Pajonal, Ucayali, Pichis, Perené,
and Apurucayali) but the scholars vary in their estimates of the number of the principal
Ashéninka dialects. Importantly, the status of Ashéninka Perené as a separate language is
not definitively ascertained. For example, Ribeiro & Wise (1978) and D. Payne (1981)
distinguish between two language varieties: one includes speakers of Perené, Pichis, and
Ucayali Ashéninka and the other speakers of Ashíninka (Axininka or Ajyininka
Apurucayali). D. Payne (1991) and Trudell (1995) include three dialect subgroupings:
Asháninka, Ashéninka, and Ashéninka Pajonal. Other conservative assessments of the
2
Sources use different labels for this language family: Arawak (e.g. Danielsen 2007, Michael 2008, Solís
Fonseca 2003), Arawakan (e.g. Adelaar 2004; Campbell 1997; Gordon 2005; Wise 1986, 1990), or
Maipuran/Maipurean (e.g. Campbell 1997; Kaufman 1994; Wise 1990).
15
dialect diversity are by Aikhenvald (1999) and Danielsen (2007) who tentatively consider
Perené, Pichis, and Ucayali Ashéninka dialects of Ashéninka Pajonal. Solís Fonseca
(2003) makes a distinction between two languages, Campa Asháninka and Campa
(Ashéninka) Gran Pajonal3. Adelaar & Muysken’s classification is the same; the
distinction is made between Asháninca and Ashéninca (2004). The classifications by J.
Payne (1989), Payne & Payne (2005), and Fabre (2005) list Pichis, Perené, Ucayali, and
Apurucayali as distinct language varieties. Gordon (2005) identifies Pichis, Perené,
Ucayali, Apurucayali, and Ashéninka Pajonal as separate languages. Alonzo Sutta posits
that the following ‘Ashaninka’ varieties are dialects: Pichis, Ucayali, Tambo-Ene, Perené,
Apurucayali, Pajonal, Yurúa (2010:81).
There is no clarity either as to what extent the Kampan languages are genetically
related on a higher-level grouping. Wise (1986) proposed an internal classification of
Kampan languages within a larger grouping of Preandine family4, including Amuesha,
Apuriña and Piro, and compared some of their grammatical characteristics (1986, 1990,
2002). D. Payne (1991) revised Wise’s (1986) classification and established the current
limits of the Kampan group, reaffirmed by Aikhenvald (1999) and Adelaar & Muysken
(2004). Michael’s (2008) most recent internal classification of Kampan languages revises
the one proposed by Wise (1986). It splits the languages into Southern (Matsigenka,
3
Varese states that the denomination Campa has no meaning to Asháninkas, other than having a derogatory
connotation. He hypothesizes that campa is a loan from an unidentified indigenous language of South
America. The term campa was introduced in 1685 by the Franciscan priest Biedma who supposedly spoke
Ashaninka and Panoan languages and was familiar with Tupi-Guaraní, which was a regional lingua franca
in those times. Varese speculates that Guaraní camba ‘black’ or ‘little black person’ applied to the jungle
dwellers in Bolivia might be related to the name given to the high-jungle inhabitants in Peru (2002:66).
Some language speakers, with whom I worked, assert that campa is a loan from Quechua kapuma which
means ‘unkempt’, ‘disheveled.’
4
Some linguists continue to use the term Pre-Andine (e.g. Cysouw 2007, Gordon 2005, Payne & Payne
2005; Solís Fonseca 2003,Wise 2002) to refer to Kampan languages.
16
Nanti, Nomatsigenga) and Northern (Kakinte, Asháninka, Ashéninka5) branches. Within
the Northern branch, Asháninka and Ashéninka are placed in one subgroup, as closely
related language varieties, while Kakinte is not considered part of this dialect chain.
Though there are no doubts concerning the genetic affiliation of Kampan languages,
problems still exist regarding the internal genetic relationships within the subgrouping
and possible genetic relationships with other Andean languages. The identification of the
possible genetic affinity of Kampan with other languages is confounded by the lack of
agreement between scholars on a mid-level grouping of Arawak languages. The main
problem of Kampan studies is the lack of adequate data for the languages that are
included in this group. It is not possible to define the internal divisions within the
Kampan subgrouping until more comparative and historical studies on Kampan
languages are published.
1.3 The regional history and status of Ashéninka Perené
This subsection sketches out a general historical background of the region where
Ashéninka Perené speakers currently reside and provides an assessment of the sociolinguistic status of the language at present. A caveat is due here. This presentation does
not aim to provide an anthropological or sociological analysis of Ashéninka Perené’s
society. Instead, it is largely an account of what I have seen or heard from my language
consultants.
Archeological explorations reveal that Arawak-speaking peoples have lived in the
area since 2200 AD (Gines 2001:40). For millennia, the speakers have had essentially
5
I preserve the spelling of language names as they were given by the sources.
17
unlimited access to the unique biodiversity of this ecological niche, located both at the
foot of the eastern Andes and the western fringe of the Amazonian jungle. Until the last
70-80 years, during which their ways of living have dramatically changed, the Ashéninka
Perené had mainly subsisted on game, fish, and slash-and-burn agriculture (for the
purpose of cultivating vegetable gardens). Before the establishment of intense contact
with outsiders and integration into the money-based citrus- and coffee-growing economy
in the 20th century, the Ashéninka speakers had extensively used natural resources to
hunt, fish, construct homes, make bows, arrows, machetes, fishing gear, looms, utensils,
cooking pots, fabric, adornments, etc.
In the post-Colombian times, there were successful attempts to penetrate Campa
territory by Jesuit and Franciscan priests and establish Catholic missions in the 16th-17th
century but the triumphant Campa uprising (with the rebels’ base in the mountainous
grasslands of Gran Pajonal) under the leadership of Juan Santos Atahualpa in 1742
stopped the process of colonization and Christianization and closed the Campa-controlled
region to missionaries, prospectors, and travelers for more than a century. Renewed
attempts to re-conquer the area began in the middle of the 19th century when the Peruvian
government dispatched a series of military expeditions aimed at capturing Campa Indians
and transporting them to the agricultural coast as a way of resolving the labor shortage
issue. The Peruvian state also directed the immigration of foreign workers to the high
jungle area and granted them ownership of the tribal lands (Varese 2002:113). The armed
resistance of Upper Perené Indians to the state-sanctioned invasion initially slowed down
the whites’ penetration but the 1868-1869 military raids during which the soldiers
completely destroyed Upper Perené Indian settlements put a halt to any significant
18
indigenous attempts to stop the aggression. Accounts of those raids narrated in Spanish
by the descendents of the warriors have been collected by the historian of Pampa Michi
Raul Martin Bernata. According to Raul Martin Bernata’s testimony, the ancestors’
response to the colonization was to flee to the surrounding mountains or fight.
Indeed Upper Perené Indians were reported to constantly clash with Italian-born
immigrant settlers (Varese 2002:132). Figure 2 is the Spanish transcript of the story about
an Indian skirmish with colonists. The transcript is displayed in the Pampa Michi
community archive building managed by Raul Martin Bernata, shown in Figure 1.
The English translation is as follows. The story that we were told is that one day our
fellowmen were surprised by a group of colonists who looked awful: they had long
beards, were wearing different clothes, their shoes were covered with dirt, and they were
yelling,’ Chuncho, chuncho, chuncho!6’ The colonists were pointing with their fingers at
our fellowmen. Our fellowmen let out the war call, ‘Ez-zih, heyuh, yeyuh, yeyih!’ Some
fellowmen responded to this call from their homes, others from the trails, and quickly
blew the horns made of snail shells, ‘Tu-u-u-uh!T-u-u-uh! Tu-u-u-un!’ It caused despair
and panic among colonists. The war calls didn’t cease: ‘Hey-ti, erah, herahz, heyi!’ The
horns responded from the hills: ‘Tu-u-uh! Tu-u-uh! Tu-u-uh!’ And everyone with their
arrows ran after the colonists, killing them. Some of the colonists escaped and told their
fellowmen that the chunchos were very fierce. The Asháninkas were yelling, ‘You are
devils! You are bringing death, disease, and bad luck to us!’ The colonists, seeing that
many Indians were running towards them, in panic retreated to their trails screaming,
‘The chunchos are coming to kill us! Chunchos! Are coming! Coming! Coming!’ They
6
‘Chuncho’ is a derogatory term used by outsiders to call indigenous high jungle dwellers.
19
were saying this to each other. That’s true that those colonists who could not run were
left behind. They were shot by arrows and killed. This happened long time ago.
FIGURE 1. The Pampa Michi historian Raul Martin Bernata gives a tour of the community archive
building
FIGURE 2.Transcript of a story about a clash of Upper Perené Indians with colonists
The late 19th -century rubber boom (which quickly ended in the early 20th century)
brought more changes to the traditional Indian lifestyle when Indian males were recruited
by rubber dealers to harvest shiringa (rubber) trees in the jungle. Even after the boom
20
ended, the rubber industry continued to undermine family and community life in the area.
The 68-year-old language consultant Bertha Rodriguez de Caleb from Bajo Marankiari
recalls that her father and later her husband would leave the family for many months to
fell rubber trees in the jungle, as cited in (1.21).
1.21
Noime te isaiki, ihatzi yantavaitzi, itotatzi shiringa.
no=ime
te
i=saik-e
i=ha-tz-i
1SG.poss=husband NEG.REAL 3m.S=be.at-IRR 3m.S=go-EP-REAL
y=aNt-a-vai-tz-i
i=tot-atz-i
shiringa
3m.S=work-EP-DUR-EP-REAL 3m.S=cut-PROG-REAL rubber.tree
‘My husband wasn’t home, he went to work, he was cutting rubber trees.’
In the 20th century the colonization of the high-jungle tribal area continued as masses
of coastal and Andean colonists arrived. The usurped lands were largely used for coffee
and citrus plantations. At this time, the British-owned The Peruvian Corporation Ltd.
received, as a debt payment, a parcel of 500,000 hectares from the Peruvian government
in the Upper Perené River area, called Colonia del Perené, with the objective of
cultivating extensive crops. The same consultant Bertha Rodriguez de Caleb (Bajo
Marankiari) recalls that after the Peruvian Corporation appropriated her village’s tribal
lands, her fellowmen planted beans and other crops in what used to be their land, but the
corporation employees would bring in a herd of cows which would destroy their gardens,
as described in (1.22-1.23).
1.22
Osaikatzika compañia, te ishinitero ante avanipaye intatzikero, te ishinitero ante
avani anpankite parentsi, kaniri.
o=saik-atz-i=ka
compañia te
i=shini-t-e=ro
3n.m.S=be.at-PROG-REAL=ADVcompany NEG.REAL 3m.S=allow-EP-IRR=3n.m.O
aNt-e
a=vani-paye
intatzikero te
make-IRR 1PL.poss=garden-PL other.side
i=shini-t-e=ro
NEG.REAL 3m.A=allow-EP-IRR=3n.m.O
21
aNt-e
a=vani
a=N-paNki-t-e
parentzi kaniri
make-IRR 1PL.poss=garden 1PL.S=IRR-plant-EP-IRR plantain manioc
‘There was a company here, they didn’t allow to make gardens on the other side
of the river, they didn’t allow us to make our gardens to plant plantains and
manioc.’
1.23
Arika inyakero apankite kaniri yamakeri ivaakate, itsonkiro yovaro maaroni
avakayariririmi neentsitpatsaini.
arika i=ny-ak-e=ro
a=paNki-t-e
kaniri
when 3m.A=see-PRF-IRR=3n.m.O 1PL.S=plant-EP-IRR manioc
y=am-ak-e=ri
i=vaaka-te
3m.A=bring-PRF-IRR=3m.O 3m.poss=cow-poss
y=ov-a=ro
i=tsoNk-i=ro
3m.A=finish-REAL=3n.m.O
maaroni a=v-aka-ia=ri=ri=mi
3m.A=eat-REAL=3n.m.O all
1PL.A=eat-CAUS-IRR=3m.O=REL=CNT.F
n=eentsi-te-patsaini
1SG.poss=child-poss-DIM
‘When they saw us plant manioc, they would bring their cows, they would eat all
that we would have fed to our little children.’
Later the corporation divided the land into lots and sold them to Andean colonists.
The capillary influx of colonists from the cordillera eventually has resulted in the
dramatic change of the consultant’s village’s make up. According to the former chief of
Bajo Marankiari Gerardo Castro Manuela, the village, founded in 1938, used to be a
native community; currently, out of one hundred thirty families only ten are Asháninka.
Based on the language consultants’ reports and my own observations, farming is
presently the main occupation of the native population (see Figure 3). The main
cultivated crops are plantains, papaya, pineapple, and citrus fruit. (See Appendix B for
description of activities in which the story-teller is engaged. ) Hunting is no longer a
viable source of food supply due to the scarcity of game while fishing and snail
harvesting yet remains a regular activity of older men and women. Ashéninkas are skilled
artisans who fabricate exquisite adornments from a variety of wild seeds for both men
22
and women. The tradition of fabricating things has continued merely in a handful of the
remaining native communities and is essentially in the state of decline. Only some crafts
have survived, evidently, due to their market value. Crafts sales, popular with tourists
from Lima, generate supplementary income to many older women in the native
community of Pampa Michi, one of the main local tourist destinations. A picture of a
craft stand is given in Figure 4. Craft-making businesses in other native communities do
not get as much tourist traffic; as a result, they are barely surviving.
FIGURE 3. A hill slope with chacras (private plots of land), with rows of fruit-bearing trees, in the
Perené District
FIGURE 4. A craft stand in Pampa Michi
23
Although Asháninkas of the Upper Perené River were able to survive the
government-supported ethnocide and have adapted to the market economy, having been
among the first jungle populations to have contact with whites, they are on the brink of
losing their mother tongue.
My daily interactions with villagers have revealed that they are well-functioning
bilinguals or Spanish monolinguals. Very few speakers, mostly elderly men and women
over 80 (some of them are centenarians), speak exclusively Ashéninka. Spanish is used in
all domains: school, local government, church, and business since buyers of Ashéninka
products or those who sell goods to Ashéninka typically speak Spanish. As the result of
permanent contact with Spanish-speaking outsiders, a majority of Ashéninka Perené has
assimilated into mainstream Peruvian society, i.e. they have acquired many aspects of
western culture (they call themselves civilizados ‘civilized’) and are fluent speakers of
Spanish. My fieldwork observations corroborate assessments of the speakers’ shift to
Spanish made by Anderson (2000:44) and Wise (1985:196, 203).
The most troubling fact is that Spanish has also become the home language. The
youngest Ashéninka-speaking parental generation (they are in their late 40s) prefer to use
Spanish as the home language and admit that they raise their younger children
monolingual, as seen in Gerardo Castro Manuela’s remark in (1.24).
24
1.24
Anyaaninta ayovetaro anyaani, intaani atomipaye te iyotahero inyaavaitahero
anyaani.
a=nyaa-ni=Nta
a=yo-ve-t-a=ro
a=nyaa-ni
1PL.poss=word-poss=DEM 1PL.A=know-FRUS-EP-REAL=3n.m.O 1PL.poss=word-poss
intaani a=tomi-paye
only
te
i=yo-t-ah-e=ro
1PL.poss=son-PL NEG.REAL 3m.A=know-EP-REGR-IRR=3n.m.O
i=nyaa-vai-t-ah-e=ro
a=nyaa-ni
3m.A=speak-DUR-EP-REGR-IRR=3n.m.O 1PL.poss=word-poss
‘Our language, we know our language, but our sons don’t know how to speak
our language.’
On average, there are six to eight children in such families. The oldest that are in their
late 20s-early 30s usually have some degree of proficiency in the language; younger
children are essentially monolingual.
Another contributing factor is speakers’ attitudes. The indigenous mother tongue has
little value in a market-oriented society and is not helpful in everyday economic struggle
to bring food home. As one speaker said, reflecting the views of her fellowmen, ‘The
language doesn’t sell.’ In the eyes of the speakers, the language is not emblematic of their
‘civilized’ identity. Sadly, it fosters the profound feeling of inferiority since the Peruvian
model of patriotism favors supremacy of the Castilian language and culture at the cost of
indigenous linguistic and cultural legacy. The current national subtractive intercultural
bilingual policy and the regional policy of indigenous language standardization largely
work against the many speakers’ needs to reclaim the oral vernacular language (Chiroque
Chunga & Rodríguez Torres 2008: 21, 37-38; Hornberger 2000:179-180; Nieroda
2005:80).
Local linguists from San Marcos University in Lima have been active in launching
literacy programs in Ashéninka varieties (Alonzo Sutta 2010) but such literacy efforts are
25
stalled by orthographies that do not reflect the phonology of the language. For example,
at the September 2010 meeting speakers from Bajo Marankiari and Pampa Michi rejected
the currently promoted standardized alphabet for Ashéninka varieties on the grounds that
it is not suitable for the Upper Perené area.
Crucially, the Ashéninka Perené comprise one of the smallest groups among the
speakers of other Kampan languages, e.g. 23,000 speakers of Asháninka, 12,000 of
Ashéninka Pajonal, 12,000 of Ashéninka Pichis, 13,000 of Ashéninka Ucayali, etc.
(Gordon 2005). Speakers of Ashéninka Perené are vastly outnumbered by Spanishspeaking colonists who reside in the Upper Perené River area. There is a high rate of
intermarriage between mestizos and Ashéninka families with Spanish used as a common
language in many social contexts such as family parties, soccer and volleyball matches.
The vitality of Ashéninka Perené can be evaluated based on the Whaley and Grenoble
diagnostics (2006:18). Whaley and Grenoble propose five levels of language vitality:
safe, at risk, disappearing, moribund, nearly extinct, and extinct. They argue that when a
language is ‘safe’, all generations use the language in all or nearly all domains, the
language has a large speaker base, and it functions as the language of government,
education, and commerce. In contrast, when a language is ‘at risk’, it’s spoken in a
limited number of domains and has a smaller number of speakers than other languages in
the same region. To be categorized as ‘disappearing’, the language should show a pattern
of the shrinking speaker base when a shift to another language is attested in the
communities where it is spoken; it’s also used in more restricted domains, and the
dominant language begins to replace it at a greater percentage at home. Based on these
metrics, Ashéninka Perené appears to be the disappearing type.
26
The Krauss framework for classifying languages according to degree of their viability
involves children, whether they learn their parents’ language or not, and includes three
major language categories: safe, endangered, and extinct (2007:1). Endangered languages
are subdivided into stable, partly stable, definitively endangered, and severely
endangered depending on the degree of brokenness of the language transmission
mechanism. Ashéninka Perené belongs to the definitively endangered language type
matching the profile described by Krauss: “the language has passed the crucial basic
threshold of viability, is no longer being learned as mother tongue by children in the
home, … the younger speakers are of the parental generation” (2007:5). Given the small
size of the surviving speakers and the intergenerational language transmission break,
Ashéninka Perené is in serious danger of extinction within one or two generations.
1.4
Prior research on the language
In this section, the name of the language and its self-denomination is discussed, along
with Ashéninka Perené scholarship. The speakers refer to themselves as katoNkosatzi
(katoNko is ‘upriver’ and –satzi is a productive derivational suffix used in names of
peoples) ‘upriver people’ or parenisatzi (pareni ‘river’ and –satzi ‘human provenance’)
‘river people’. Another autodenomination is ashaniNka or asheniNka a=sheni-Nka
[1PL.poss=be.of.the.same.group-NMZ] ‘our fellowman’7. The term ashaniNka/
asheniNka usually bears an accent mark on the second vowel when its hispanized version
appears in print. The stem sheniNka can be marked for varied possessors, e.g.
7
Jacinto Santos contends that the origin of the term asháninka/ashéninka lies in the speakers’
conceptualization of the world. The essence of his hypothesis is that asháninka/ashéninka is a nominal
compound consisting of ash ‘four worlds’ and ninka ‘interrogative pronoun who’; the figurative translation
of asháninka/ashéninka should be ‘be the protector of’ (2010: 68-69).
27
nosheniNka ‘my fellowman’, pisheniNka ‘your fellowman’, osheniNka ‘her fellowman’,
etc. When speakers use Spanish to describe themselves, they say ‘Asháninka del Alto
Perené’, i.e. Asháninka from the Upper Perené River. The language is referred to by the
speakers as anyaani katoNkosatzi parenini a=nyaa-ni katoNko-satzi parenini
[1PL.poss=word-poss upriver-provenance river] ‘our word of the Upper River’. In
linguistic scholarship, various terms are used for the speakers’ denomination and the
language’s name: Upper Perené (Payne 1981:5), Upper Perené (Campbell 1997:181),
Perené, dialect of Pajonal Campa (Aikhenvald 1999:68), Alto Perené (Heise, Payne,
Payne, Vilchez 2000:5), Campa Asháninka (Solís Fonseca 2003:144), Perené Ashéninka
(Gordon 2005), Campa Alto Perené (Fabre 2005:31).
Limited accounts of grammatical features of Ashéninka Perené are found in
publications by SIL and local linguists. An illustrated dictionary of five Ashéninka
varieties published for the local bilingual teachers contains a small number of Perené
words (Heise et al. 2000). There is a short survey of pronominal forms in three
Ashéninka varieties, Pichis, Perené, and Apurucayali (Reed & Payne 1986). J. Payne’s
useful pedagogical manual (1989), intended for the beginning students of Ashéninka
Perené and Pichis, provides a brief account of the pronominal system, possessive
prefixes, TAM, negation, complements, directionals, applicative and reflexive suffixes. It
contains samples of connected speech such as dialogs and folk stories, along with
comments on the orthographic conventions, sound system, stress patterns, and rules of
vowel elision. There are good quality accounts of phonology and morphology of
Ashéninka Pichis (Payne 1983), Ucayali (Garcia Salazar 1993), and Apurucayali (Payne
1981; Payne, Payne & Sanchez Santos 1982; Spring 1992; Black 1991) but no
28
comparable work on Ashéninka Perené except for an outline of lenition and elision by
Falcon (1993). The Ashéninka speakers’ pragmatic considerations in the choice of verb
aspect are addressed in Payne (1984). Payne & Payne (2002) discuss the development of
causative constructions and split intransitivity in five Ashéninka languages including
Perené (2005). Cysouw (2007) considers briefly the origin of interrogative words in
Perené in the context of his discussion of content interrogatives in Pichis Ashéninka.
Alonso Sutta (2007) is an investigation of the classifier –tha in Ashéninka Perené. Jacinto
Santos (2010) is a hypothesis on the etymology of the word ‘asháninka’. General sociolinguistic information about the language and its cultural context is provided in Anderson
(2000, 2005), Weiss (1974, 1975), and Wise (1985).
1.5
Fieldwork materials, research methods, and consultants
The grammar is based on the materials collected during three fieldtrips in June 2008
(1-week familiarization trip), June-September 2009 (12-week trip), and May-September
2010 (16-week trip). The collected corpus is comprised of 18 hours of video recordings,
22 hours of audio recordings, 387 pages of notes, vocabulary entries, and comments on
collected texts taken during meetings with language speakers, 236 pages of annotated
digital texts, and 468 pages of transcribed and translated texts performed by the primary
language consultants Gregorio Santos Pérez, Daniel Bernalez Quillatupa, and Delia
Rosas Rodríguez. The fieldwork was conducted in Distritos Perené and Chanchamayo,
Provincia Chanchamayo, Departamento Junín. The only way from Lima to
Chanchamayo Province is by bus which involves a 7-9 hour ride (depending on the traffic
and weather) across the Andes. The Perené District is located 330 km to the east of Lima,
29
the capital of Peru. The District’s lowest altitude is 505 meters above the sea level, the
highest is 3,300 meters.
I was based in La Merced, the capital of Chanchamayo Province and traveled daily to
the villages of Pampa Michi (18 km from La Merced), Santa Ana (25 km from La
Merced), and Bajo Marankiari (26 km from La Merced) to meet with my consultants.
Recruitment of consultants was initially a challenge. The outcome of my inquiries with
the indigenous political organization CECONSEC (La Central de Comunidades Nativas de
la Selva Central) was unsuccessful when I petitioned them to give me permission to
travel freely to local native communities to recruit language consultants. In June 2009,
the leader of CECONSEC Perry Amato Bonatto assessed the project as ‘unnecessary’ and
warned me not to enter communities on my own. This cautious stance might have been
triggered by the violent political disturbances in June-July 2009 when indigenous people
across the country were participating in mobilizaciones (demonstrations, strikes, and road
blockades) as a form of protest against the pro-western Alan García government.
Since I didn’t have permission from CECONSEC to work in native communities, my
best chance was to establish private contacts with native speakers through the
acquaintances that I had in La Merced. Aldo Sedano Marro, manager of a small private
English school in La Merced, was instrumental in the recruitment of my first language
consultant Raul Martin Bernata from Pampa Michi. Through Raul I met Gregorio Santos
Pérez, a native of Bajo Marankiari, currently residing in Villa Perené. These two have
become my principal language consultants. They gradually engaged their families and
canvassed their friends to solicit their participation in language documentation. Another
useful contact was Lorena Tremolada Broque, manager of the fundo San Jose where I
30
was staying, who introduced me to Daniel Bernales Quillatupa, a bilingual specialist from
the local Department of Education, also a native of Bajo Marankiari. At the request of
Gregorio Santos Pérez, the chief of Bajo Marankiari Osbaldo Rosas Rodriguez granted
me permission to work in the community. As the result, my main research community
became the village of Bajo Marankiari, thanks to the welcoming reception of its chief
who unconditionally embraced the language documentation and grammar-writing project.
Although I wasn’t able to visit other villages (except for San Miguel Centro Marankiari
which I visited in June 2008), I obtained audio recordings of speakers from the villages
other than Pampa Michi, Santa Ana, and Bajo Marankiari, collected at my request by the
language consultants Daniel Bernales Quillatupa and Gregorio Santos Pérez. The
sampled communities include Churingaveni, Mariscal Cáceres, Santari, Alto Esperanza,
Pucharini, Santo Domingo, and Platanillo Shimaki.
The main methods of data collection for the language documentation and grammarwriting project included participant-observation, collection of oral and written texts, and
direct elicitation, with audio and video recordings of different registers of connected
discourse (narrative, conversation, song, dance), along with linguistic analysis of the
collected data. Each consultant was paid hourly at the end of our meetings. Our meetings
took place during the time when consultants were available, which was usually from
8:30-9:00 in the morning until 4:00-5:00 in the evening. I would normally work with 3-5
language consultants daily. The consultant questionnaire was administered to obtain
metadata to accompany the collected recordings. The vocabulary was elicited based on
the Comrie & Smith basic wordlist (1977), SIL Comparative African Wordlist
(SILCAWL) (Snider & Roberts, http://www.sil.org), Comparative Kampan vocabulary
31
list (Michael ms.), and rechecked against D. Payne (1980), J. Payne (1989), and Heise et
al. (2000). Some grammatical information was elicited based on Comrie & Smith (1977),
Payne (1997), and Dixon (2010a, 2010b) and analyzed via a large body of naturally
occurring texts. Our meetings were recorded using a Sony PCM D-50 Linear PCM digital
recorder, a Panasonic HDC-HC100P/PC digital camcorder and Sony Electret Condenser
Stereo Microphone ECM-MS908C, with the elicited audio and video data transferred to
the laptops Dell B130, Dell Mini, Dell Latitude E6410, and WD Western Digital ‘My
Passport Essential’ portable hard drive. The transferred digital files were edited using the
Audacity Sound Editor Software (http://audacity.sourceforge.net/) and HD Writer 2.6E
High Definition Image Management/Easy Editing Software. Lexical items were
transcribed, analyzed, and entered into an Excel spreadsheet. I utilized the PRAAT
software application (http://www.fon.hum.uva.nl/praat), JPlotFormants v1.4 Formantplotting Software (http://www.linguistics.ucla.edu/people/grads/billerey/PlotFrog.htm ),
Ladefoged (2001, 2003, 2005), and Ladefoged & Maddieson (1996) for speech analysis.
Some of the video recordings were accompanied with annotations using the ELAN
software application (from MPI-Nijmegen). The recorded texts were transcribed by the
data provider and translated by me and/or by the data provider. The accuracy of
transcriptions and translations were verified by other consultants. The language
consultants who participated in collection and linguistic analysis of the data were trained
individually during the initial phase of the project. The training was conducted at the
consultants’ home residences, Internet outlets in Santa Ana, and at Instituto Cultural
Peruano Norteamericano Region Centro (IPCNA) in La Merced. Primary consultants
were trained to work with digital recorders, the video camcorder, and the Dell Mini
32
laptop. Technical support and training was provided to the consultants throughout the
project to ensure good quality of the documented materials.
During the first break from the field work in 2009, I performed a systematic linguistic
analysis of the collected data, revised the preliminary grammatical descriptions made in
the field, and identified the areas in need of more research. In 2010, I engaged the team of
consultants with whom I worked in 2009 and continued to train the consultants in
language documentation and analysis, verify and enter the collected data into data files,
and perform a linguistic analysis of the data. After a second field trip to the area in 2010,
I completed the analysis of the collected materials and finished the grammar sketch,
which is my doctoral dissertation. Data providers, Centro Amazonico de Antropología y
Aplicación Practica (CAAP) in La Merced, and Instituto de Investigación de Lingüística
Applicada (CILA) in Lima, Perú were given digital copies of audio and video recordings.
A digital collection of Ashéninka Perené texts was archived in 2009 at the Archive of
Indigenous Languages of Latin America (AILLA), University of Texas at Austin
(www.ailla.utexas.org).
From the start of the language documentation and grammar-writing project, the
speakers asked about the practical outcome of our team work and urged me to consider
using some of the grant money on the publication of a storybook for the community,
based on the collected recordings. The literate speakers’ main sentiment was
dissatisfaction with the substance of school materials disseminated by the local
educational authorities. Although the native language is supposed to be taught as a
foreign language in the District’s elementary schools, the orthography, grammar, and
vocabulary of school texts do not represent the Ashéninka Perené variety. The problem is
33
that following the current policy of standardization of the Ashaninka/Ashéninka varieties
(Pajonal, Ucayali-Yurua, Apurucayali, Tambo-Ene, Pichis, and Perené), local educational
authorities promote as the standard the numerically most dominant Tambo-Ene variety.
To accommodate native speakers’ request, the language consultants’ team decided to
focus on the development of orthography and selection of stories for the proposed
storybook. The team also suggested to include genres other than stories such as songs,
riddles, tongue-twisters, explanations of beliefs and customs, and traditional advice given
by the parents to the children.
During the summer 2010, language consultants were mainly involved in the
preparation of texts for publication. Their duties included recording fellow speakers and
themselves, transcribing and translating collected recordings, recruiting fellow speakers
as contributors to the project, reading and editing the printed copies of the selected texts,
and making decisions about the practical writing conventions. The three principal
language consultants collected texts (traditional and children’s stories, accounts of
everyday activities and historical or social events, conversations, commentaries on beliefs
and customs, songs) with the objective of recording 8 hours of connected discourse using
the following audio and video equipment: a Panasonic HDC-HC100P/PC digital
camcorder, two Tascam DR-7 digital recorders, and an Olympus WS-320M digital voice
recorder. The collected texts were transcribed by a language consultant other than a
narrator and translated. The accuracy of transcriptions and translations were rechecked, at
least once, by other speakers. The preliminary draft of the storybook was approved at the
September 2010 meeting of language consultants in Bajo Marankiari. The goal is to
34
disseminate 130 copies of the published storybook to the District’s bilingual teachers,
language consultants’ households, and to more remote language communities.
Language consultants who actively participated in the project included nine women
and nine men. From those, the youngest consultant Jakeline Magbel Castro is 29 (born in
1982), the oldest are Ruth Quillatupa Lopez, 69, Bertha Rodriguez de Caleb, 69 (born
1941), and Abdias Caleb Quinchori, 68 (born 1942). The consultants belong to 5
families: the Martin family in Pampa Michi, and the Castro/Rosas, Santos/Pérez, Caleb,
Quillatupa, and García families from Bajo Marankiari. Five consultants have good
education; three are college-trained bilingual teachers, two finished high school. Others
have either basic literacy skills or are illiterate. Consultants from Bajo Marankiari are
members of the village Seventh Day Adventist church which was founded by the SDA
missionary Frederick (Fernando) Stahl8.
GREGORIO SANTOS PÉREZ, 50, a bilingual teacher trained by SIL missionaries, was
my main consultant. He is a spectacular speaker of Ashéninka Perené and an insightful
analyst of his language’s grammar intricacies. Gregorio Santos is an outstanding narrator
who uses elaborate complex sentences, falling and rising sentence cadences, ideophones,
facial expressions, and gestures to convey a story. He was raised in Bajo Marankiari
where his parents are still residing and moved to Santa Ana fifteen years ago. Since he
was educated and worked for a while in the area populated by Ashéninka Pichis, he
sometimes uses non-Perené vocabulary items, which draws criticism from his fellow
8
Language consultants report that the first Adventist mission was originally founded in Metraro (Metaro)
in 1921. Many natives from the Upper Perené area became Adventists because they believed that Stahl was
Pava, the solar deity, who arrived to save them. Despite the fact that Stahl left the area in 1928, the
Adventist movement spread along the Perené and Tambo rivers. Bajo Marankiari was founded by Indian
Adventist believers. A detailed account of the mass conversion of Upper Perené natives to Adventism,
Catholicism, and Evangelism during the colonial and postcolonial periods is Santos-Granero (2002, 2004).
35
speakers. Gregorio Santos has proven to be a charismatic leader, an organizing force
within the language consultants’ team and native community, and is genuinely concerned
about the language’s prospects. In 2010, at my request he traveled to remote communities
to record and write down speakers’ narratives. Gregorio Santos also independently
worked with the SIL African Comparative Wordlist and translated over 500 vocabulary
items. Gregorio Santos is the only consultant with whom I stay in contact via telephone
and email during the breaks from fieldwork. Gregorio’s wife, Dora, uses Spanish as the
daily medium of communication with her family and friends, but her passive knowledge
of the language helped me evaluate sample school texts used by bilingual teachers in
elementary schools to pinpoint the differences between the Ashéninka and Tambo-Ene
varieties. Sadly, only two (the oldest) of his eight children speak the language. One of
them, Edgardo Santos Meza, is a bilingual teacher. I got a chance to work with him once.
During that meeting, Edgardo made illuminating comments about the plural markers in
the language and the polyfunctionality of the form ni.
DELIA ROSAS RODRIGUEZ, 41, is a very proficient speaker and a tireless transcriber
of collected texts. Delia was born in Bajo Marankiari and most of her life resided in the
village. As a teenager, she moved to Lima in search of work and worked as a house maid
for a number of years. She finished high school and did one year of college coursework
in Lima but lack of sufficient means to continue education derailed her ambition and
forced her to come back home. Delia Rosas is a language purist and rejects innovations or
deviations from the ‘norm’ as ‘incorrect’ or ‘unnecessary’. Delia considers her mother
Bertha Rodriguez de Caleb an authority on language issues and consults her when in
doubt. When asked to translate vocabulary lists, Delia chose to do the work with Bertha.
36
Delia daily speaks the language with her mother, stepfather Abdias Caleb, and sister
Victorina Castro. Delia has good Spanish literacy skills and is a quick study. In 2010, she
learned how to use the digital recorder, video camera, and laptop and, at my request, did
video and audio recordings of seven Ashéninka-speaking villagers and transcribed and
translated most of the recorded texts. Delia’s embroidery work was an inspiration for
recording a cycle of stories about Ashéninka warriors when her tablecloth with the
embroidered images of warriors was used as a prompt for the narrators. Delia has two
daughters; one of them, Jakeline Magbel Castro, speaks the language.
The daughter of Delia Rosas JAKELINE MAGBEL CASTRO, 29, was raised by her
paternal grandmother Dominga, who was a speaker of Ashéninka Pichis. Jakeline
remembers her grandmother’s stories, and is an enthusiastic story-teller and singer, but
her speech is replete with non-Upper Perené words and structures. Delia and Gregorio
had to double-check and correct the data that she provided. Jakeline has a knack for
drawing and drew wonderful illustrations for the storybook’s riddles.
DANIEL BERNALES QUILLATUPA (Aroshi), 48, is a very proficient speaker and one
of the main contributors to the language documentation and grammar-writing project. As
a bilingual specialist with the local Department of Education, he oversees the work of
bilingual teachers in Chanchamayo Province. Born and raised in Bajo Marankiari,
educated by SIL missionaries, Daniel has excellent Spanish and Ashéninka literacy skills
and a vast knowledge of Ashéninka culture. Because he was away on business trips four
days a week, he chose to work independently, recording other speakers and transcribing
and translating the recorded texts. Daniel is also a talented artist who drew illustrations
for the forthcoming Ashéninka Perené storybook. Daniel supports the standardized
37
Asháninka alphabet on the grounds that the Asháninka/Ashéninka varieties are closely
related and there is no point in inventing a separate set of writing conventions for each of
them. Daniel’s wife and children do not speak the language. In the home, the only family
member with whom he speaks the language is his mother Ruth Quillatupa Lopez.
The mother of Daniel Bernales Quillatupa, RUTH QUILLATUPA LOPEZ, 69, is a
talented narrator who contributed the highest number of stories. The language consultants
who did edits of the collected stories commented that her stories strike imagination, are
marked by ornate detail, and always have an unusual twist. Ruth is very articulate, clear,
and grammatically accurate. The aspirated /tʰ/ before /o/, e.g. irithori ‘his sister’, thonk
‘finish’ makes her speech slightly conservative.
RAUL MARTIN BERNATA, 58, was my first teacher of Ashéninka Perené. Raul
Martin has lived most of his life in Pampa Michi. Most of Raul’s interactions with the
family and villagers are in Spanish. As a retired paramedic, he possesses an advanced
knowledge of medicinal plants and is one of the best consultants to talk to about
Ashéninka vocabulary. Raul knows many myths and is well-versed in Ashéninka history.
In 2010, he became the guardian of the community ‘history’ building which was
constructed at his initiative. The building houses pictures, maps, stories about the history
of the community of Pampa Michi, and artifacts from the villagers’ past life,
accompanied by short texts in Spanish. Raul was instrumental in the recruitment of his
wife and his brothers-in-law for the purpose of language-documentation work.
Raul’s wife, VICTORIA MANCHI DE MARTIN, 51, is a very proficient speaker.
Victoria was born in Bajo Marankiari and moved to Pampa Michi after marrying Raul.
Victoria volunteered for short meetings to verify accuracy of transcribed texts and
38
recorded her demonstrations of artifacts production (necklaces, wrist ornaments, money
bags, fire fans, mats, etc). Victoria also worked with her brother Ernesto and husband
Raul on the compilation of detailed vocabulary lists. None of the couple’s 11 children are
competent speakers of Ashéninka Perené. Their oldest children who come to visit the
parents from the coast have a good passive knowledge of the language, but are not
confident speakers because the parents don’t speak Ashéninka to them; Spanish is the
home language. The eldest daughter Eliana (she is in her early 30s) thinks that another
reason why her younger siblings failed to learn the language is that it was never taught in
the community elementary school.
LUIS MAURICIO ROSA, 61, and ERNESTO MANCHI LOPEZ, 63, are Victoria Manchi
de Martin’s brothers. Both are very proficient speakers and fabulous story-tellers; Ernesto
also knows a lot of songs and plays well the Ashéninka flute sonkari. Luis’s speech is
distinctive in that when he talks about hypothetical events, he does not use the first
element of the discontinuous irrealis status morpheme N-…-e/-ia (neither does his
brother-in-law Raul). Ernesto’s speech is more representative of the younger generation
of speakers and is equivalent to that of Gregorio Santos. The brothers do not speak the
language to their children and grandchildren because they do not see them much. Neither
brother has a wife (Luis is a widower and Ernesto is divorced) and both are constantly on
the move in search of jobs. Their specialty is construction of houses with thatch-woven
roofs, which requires a unique skill. The brothers are in demand since the numbers of
knowledgeable builders are dwindling.
BERTHA RODRIGUEZ DE CALEB, 69, is an extremely proficient speaker and a gold
mine of cultural information about the language. Bertha is a fabulous narrator who
39
contributed dozens of texts to the 2009-2010 collection. Although her Spanish is fluent,
Bertha Rodriguez prefers to speak Ashéninka and uses it with her husband Abdias Caleb,
daughters Delia and Victorina, other adult members of her extended family, and
neighbors. Bertha was born in and lived most of her life in Bajo Marankiari, although her
puberty years were spent in a small community not far from Puerto Bermudez (the Pichis
River area). Bertha articulates the negative particle airo as /eɾo/ rather than /eiɾo/ like
other villagers, or utilizes the third person masculine subject proclitic ir= rather than y=
before the a-initial verbal stems. Villagers comment that this uncommon articulation is
the result of Bertha’s long contact with the Pichis Ashéninka variety. Another criticism is
that Bertha’s speech is sprinkled with Spanish words, especially with coordinators como
‘as, like’, pero ‘but’, and en cambio ‘in contrast’. Bertha has 6 children; three of them
live in Bajo Marankiari. The youngest child Osbaldo Castro Rosas (who is in his middle
30s) is the current chief of the village. While his older siblings Victorina and Delia are
competent speakers of Ashéninka, Osbaldo is reported to be marginally fluent. A few
times that I saw Osbaldo, I did not hear him speak the language.
Delia Rosas Rodriguez’s sister and Bertha Rodriguez de Caleb’s daughter,
VICTORINA ROSAS DE CASTRO, 49, is a marvelous speaker and a story-teller. Her
speech is elaborate, abundant in rarely-used words, syntactically complex (she uses a lot
of subordinated sentences), and equals in its clarity and elegance that of Ruth Quillatupa
Lopez. She is probably one of the best female speakers of her generation. Like her
mother and sister, Victorina articulates the negative particle airo as /eɾo/. Victorina is
very sensitive to the issue of borrowing words from Spanish. She insists that borrowing
from Spanish is a sign of incompetence in the native language. She has 6 children.
40
Victorina’s husband, GERARDO CASTRO MANUELA, 53, is a retired bilingual
teacher and a former chief of Bajo Marankiari who became involved with the languagedocumentation project at a later stage. Gerardo was on the readers’ team which checked
the accuracy of the transcribed and translated texts although he recorded a few stories of
his own. Since Gerardo’s late mother was from the Pichis area and his father is
Matsigenka, he hadn’t been a confident speaker of Ashéninka Perené until he entered the
SIL-run teacher-training college and worked for many years as a bilingual teacher in
remote Ashéninka-speaking communities. Although villagers do not criticize his speech,
when the data that he had volunteered were rechecked, a fair number of mistakes were
found and corrected by Gregorio Santos and Delia Rosas. The couple’s younger children
do not speak Ashéninka Perené. One of the couple’s oldest children (who is in his late
20s), Deiky Castro is a bilingual teacher. It is difficult to say if Deiky is a proficient
speaker because I heard him speak Ashéninka only once. Deiky is a supporter of the
native language standardization policy.
INES PÉREZ DE SANTOS, 65, is a virtuoso as a native speaker and a story-teller. Ines
is one of the best female speakers of her generation. In the narratives, Ines likes to use
direct speech and impersonate the characters she is talking about. Her language is
representative of older generations of speakers, e.g. she uses /s/ rather than /ʃ/ before /i/,
and articulates aspirated /tʰ/ before /a/. Ines avoids Spanish loans and invents ad hoc
words on the spot to name objects for which there is no word in the language, e.g. she
called the audio recorder aakotanari nonyaavaitzi ‘the one which grabs [the voice] when
I speak,’ for her a syringe is chakoritakiri ‘the one that stings’. She was born in Bajo
Marankiari and lived most of her adult life in the village. She speaks the language to her
41
husband Moises Santos Rojas, son Gregorio Santos Pérez, her other children (she has 8),
members of her extended family, and neighbors. Villagers admire her whistling skill. She
is the only one left in village who can produce calling sounds using fingers of one hand
or two hands. These sounds are used in the jungle to alert fellowmen that the caller has
lost his/her way. Ines Pérez earns her living by healing people.
Ines’s husband, MOISES SANTOS ROJAS, 66, is a proficient speaker of Ashéninka
Perené and a decent story-teller. His speech is representative of the older generation in
that he uses /s/ before /i/ where younger speakers articulate /ʃ/; he also lists naakaite ‘we
(excl)’ on the pronominal set, rejected as archaic usage by the younger generation,
including his son Gregorio.
Ines’s brother, ALBERTO PÉREZ ESPINOZA, 64, also resides in Bajo Marankiari.
Both Moises and Alberto are avid fishermen and artisans. Alberto’s level of competence
and story-telling skill are close to those of Moises.
ABDIAS CALEB QUINCHORI, 68, the husband of Bertha Rodriguez, was one of the
main contributors to the project in 2010. Abdias was elected chief of Bajo Marankiari in
the past. At present, he toils in his vegetable garden and makes brooms, fire fans, and
other household objects for sale. Although he was born and lived many years in Bajo
Marankiari, his speech is influenced by the Ashéninka Pichis variety due to a long period
of time that he had spent in the Pichis area when he was young. Another circumstance
which somewhat compromised his Ashéninka Perené speech is that his first wife was
Yanesha and he spoke Yanesha to her. Nevertheless, Abdias was one of the most
enthusiastic supporters of the storybook-production project. He is considered a competent
speaker and speaks the language daily to his wife and neighbors. His knowledge of
42
Ashéninka lexicon is amazing. Abdias served on the readers’/editors’ team and also
recorded a number of stories for the language-documentation project. His texts were
checked for mistakes by his wife Bertha Rodriguez, Paulina Garcia Ñate, and Gregorio
Santos.
Abdias’s sister, PAULINA CALEB DE LEON, 63, is a fine story-teller and a very
proficient speaker. Her speech is clear, well-paced, and grammatically accurate. Paulina
earns her living as a seamstress (she has a sewing machine) and has resided in Bajo
Marankiari all her life. Most of her interactions with the family members and villagers
are in Spanish.
PAULINA GARCIA ÑATE (in her mid-60s) is a very proficient speaker and a
spectacular singer. Her main contribution to the project was recording of songs and
explanation of occasions on which they are sung, e.g. when the sun rises or goes down, or
at a beer party, or when a girl is in love, or when a woman is dumped by her lover, etc.
Paulina also possesses an extensive knowledge of Ashéninka warfare. One of her
children is living with Paulina and her live-in boyfriend. The daughter (in the mid-30s) is
fluent in Ashéninka; Paulina’s grandchildren are Spanish monolinguals. In her
interviews, Paulina laments about the abandonment of the weekend beer parties that used
to bring the villagers together for weekend fun, dance, and song. Paulina blames
Adventist views that prohibit public alcohol intake for the decline of festive social life in
the village.
Other consultants with whom I met at least once and recorded their speech were
Pablo Jacinto Pedros (Bajo Marankiari), Marcos Antonio Santiago (La Merced), Maria
Virginia Lopez (La Merced), Elias Meza Pedro (Mazamari), Hilda Villegas (Mariscal
43
Cáceres), and Abdias Espirito Kintori (Mariscal Cáceres). In my grammar, I also used
texts, independently recordered and transcribed by my primary consultants. The data
providers for the texts collected by Daniel Bernales Quillatupa included his mother Ruth
Quillatupa Lopez, Guzman Segundo Yamane (Puerto Bermudez) and Otoniel Ramos
Rodriguez (Churingaveni). Gregorio Santos Pérez collected data from the following
language speakers: Nestor Ramirez Rojos (Santo Domingo), Cristobal Jumanga Lopez
(Pucharini), Clelia Mishari (Mariscal Cáceres), Natalia Camarena Jacinto (Alto
Esperanza), Avelina Lopez (Alto Esperanza), Edgar Antonio Manuel (Santo Domingo),
Mateo Cristobal Casanto (Platanillo Shimaki), Livia Julio (Mariscal Cáceres), Rafael
Martinez Quinticuari (Mariscal Cáceres), Amelia Santos (Alto Esperanza), Abraham
Jumanga (Pucharini). The language consultants who actively contributed to the storybook
project are seen in Figure 5.
FIGURE 5. Ashéninka Perené language consultants from Bajo Marankiari and Pampa Michi
44
Chapter 2
Phonology
This description of the phonemic inventory of Ashéninka Perené largely reflects the
speech communities of Pampa Michi and Bajo Marankiari, respectively representing the
western and eastern varieties of the language (see Map 1). Ashéninka Perené has 8
phonemic vowels and 4 phonemic diphthongs /ai/,/oi/, /ea/, /ia/. Vowel length is
phonemically distinctive but the number of such words with the contrastive vowel length
is not high. Word stress is not phonemic and is largely predictable.
This chapter will proceed as follows. The chapter begins with the discussion of the
vowel phonemes in §2.1 and the consonant inventory in §2.2 followed by the description
of sounds in Spanish loanwords in §2.3. Phonotactics is discussed in §2.4, and intonation
in §2.5. Phonological differences in Ashéninka Perené speech are outlined in §2.6, the
orthographic system and morphophonemics are addressed in §2.7 and §2.8 respectively.
2.1 Vowels
Ashéninka Perené has an 8 vowel phoneme system, as given in Table 1. It effectively
uses maximally opposed vowel places of articulation in the phonological space. Vowels
contrast in quality in terms of height (high, mid, low) and backness (front, central, back).
There are four front vowels /i i: e e:/, two central vowels /a a:/, and two back vowels
/o o:/, the last two are the only rounded vowels. The vowels occur in all environments
within the word. Vowel length is distinctive and will be indicated by the colon, according
to the IPA notation. In practical orthography, vowel length is expressed by doubling the
vowel.
45
TABLE 1. The vowel phonemes
High
Mid
Low
Front
i, i:
e, e:
Central
Back/Rounded
o, o:
a, a:
2.1.1 Short vowels
A summary of the phonetic realizations of the short vowels is provided in Table 2.
TABLE 2. Phonetic realizations of short vowels
High
Mid
Low
Front
i [ɪ,ɨ,e,ɪ̥ ]
e [e, ε, ə, e̥, ɪ]
Central
Back/Rounded
o [o,ʊ]
a [a, e, ə]
The vowel /i/ is a high front vowel which is pronounced like the vowel [ɪ] in English pit.
It has phonetic realizations [i], [ɨ], [e], and [ɪ̥ ]. The default realization of the vowel /i/ is
[ɪ] as in kovitsi [ko.̍ß̞ɪ.tsʰɨ] ‘pot’, varipa [wa.̍ɾɪ.pa] ‘chicken’, kinkivari [kɪŋ.̍kɪ.wa.ɾɪ] ‘old’.
After affricates /tsʰ/ and /ts/ and sibilants /ʃ/, /s/, it often has a surface realization of [ɨ],
e.g. netsi ‘louse’ [netsʰɨ]. In fast speech, when the weak unstressed /i/ is found after
voiceless consonants, it can be elided or may take the phonetic value of a devoiced high
front vowel [ɪ̥ ], as in shima [ʃɪ̥ ma] ‘fish’, tsika [ʦʰɪ̥ ka] ‘where’, shintsipaa [ʃ̥ɨntsʰɪ̥ pa:]
‘raft’, kapicheeni [kapɪ̥ ʧeene] ‘little’. The unstressed phoneme /i/ when found wordfinally can take the phonetic value of [e] as in ivenki [i.̍ß̞εŋ.ke] ‘his magic plant’; in this
case, the sound cannot be distinguished from /e/, which sometimes makes it difficult to
ascertain which phoneme is the underlying one. The surface realization of /i/ as a
devoiced high front vowel [ɪ̥ ] is illustrated in Figure 6.
46
FIGURE 6. The devoiced high front vowel [ɪ̥ ] in kapicheeni ‘little’
k
a
p
ɪ̥
e:
ʧ
n
e̥
The mid-front vowel /e/ is pronounced like the cardinal vowel [e] in Spanish mesa
‘table’. It is realized more open in stressed syllables, coming close to the open-mid [ε].
The phoneme /e/ is always more open than /i/ but the absolute height of this vowel is
speaker-dependent; in many words, variation between the close-mid /e/ and open-mid /ε/
is observed among the speakers. Figure 7 shows the surface realization of the phoneme
/e/ in sheteni [ˈʃεdenε] ‘afternoon’.
FIGURE 7. Medial and word-final realization of the phoneme /e/ in sheteni ‘afternoon’
ʃ
ε
d
e
n
e
In unstressed syllables, the phonetic value of /e/ ranges from [e], to [ε], to [ə], to [e̥].
The allophones [e] and [ə] are in free variation when they are found in unstressed
‘extrametrical’ syllables as in a.̩vi.shi.̍mo.te.<me.ni> [a.̩ß̞ɪ.ʃɨ.̍mo.te.mə.ne] ‘that may
happen’. The extrametrical syllables which do not participate in stress assignment are
shown in angle brackets. The vowel /e/ has a tendency to become devoiced in casual
47
speech when found in unstressed syllables between voiceless consonants, e.g. tekatsi
‘negative existential’ [te̥.̍ka.tsʰɨ] or word-finally, e.g. kapicheeni [ka.pɪ̥ .̍ʧe:.ne̥] ‘little’, or
it may be completely elided.
Note that the mid-front vowel /e/ is typically realized higher, as [ɪ] in the westerly
speech communities of Ashéninka Perené. For example, westerly speakers say ivenki ‘his
magic plant’ as [ɪß̞ɪŋke], kireeki ‘money’ as [kiɾɪɪke], shaveta ‘butterfly’ as [ʃaß̞ɪta]
(details on dialectal phonological differences are provided in §2.6). In its surface
realization, /e/ varies across speech communities; the westerly speech community of
Pampa Michi tends to articulate this phoneme as the high front [ɪ]9. The formant plots of
the four short Ashéninka Perené vowels articulated by the female and male speakers from
Pampa Michi (the westerly variety) and by the female and male speakers from Bajo
Marankiari (easterly variety) are provided below in Figures 8-9. The frequency of the
first formant is plotted on the ordinate (vertical axis), and the second formant is plotted
on the abscissa (horizontal axis). The vowels were recorded in words pakitsa ‘hawk sp.’,
oketsi ‘eye’, kotsiro ‘knife’, tekatsi ‘negative existential’. As Figure 8 shows, the acoustic
dimensions of the two front vowels /ɪ/ and /e/ as produced by the two speakers from
Pampa Michi are very similar: the measurements of F1 are nearly overlapping whereas
the values of F2 have slight differences. In Figure 9, the acoustic record illustrates a more
significant divergence between the F1 values in the front vowels /ɪ/ and /e/ as produced
9
This analysis of the Ashéninka Perené vowel system contradicts some statements in Kampanist literature
about the presumably minimalist Ashéninka Perené three-vowel system /iao/ (e.g. Payne 1989:21).
Speakers of the easterly speech communities in the Perené District are especially conscious of the contrast
between /i/ and /e/ and make comments about the ‘vulgarity’ of the westerly speech because of its higher
articulation of /e/, close to [ɪ].
48
by the two speakers from Bajo Marankiari. In general, measurements of the front vowels
articulated by other speakers from these two communities exhibit relatively similar
results, i.e. the F1 values of /ɪ/ and /e/ nearly match when pronounced by speakers from
Pampa Michi and differ when articulated by speakers from Bajo Marankiari.
FIGURE 8. Formant plots of Ashéninka Perené vowels of the male and female speakers respectively from
Pampa Michi (PM=westerly speech variety)
FIGURE 9. Formant plots of Ashéninka Perené vowels of the male and female speakers respectively from
Bajo Marankiari (BM=easterly speech variety)
The following presents near-minimal pairs for /i/ and /e/.
2.1
ivenki ‘his magic plant’
shaveta ‘butterfly’
netsi ‘louse’
ovinkiri ‘she jumped into the river’
pivito ‘your canoe’
pinitsi ‘magic plant’
49
The open low central unrounded vowel /a/ has three phonetic realizations, [a], [e], and
[ə].The realization of [e] occurs in casual speech when the first element of the diphthong
/ai/ is raised to the close-mid front vowel [e]. As (2.2) shows, the raising of /a/ to [e]
happens when the diphthong is followed by the clitics –ri/-ro.
2.2
iyomitairi [ɪjomɪteʲɾɪ] ‘he teaches them’
airo [eʲro] ‘negative irrealis marker’
ivairo [ɪß̞eʲro] ‘his name’
itsipatairo [ɪtsʰipateʲro] ‘he combines
it’
When unstressed, the phoneme /a/ is sometimes realized as schwa [ə], as in tsika paita
[tsʰɪ̥ kə paʲta] ‘what’. Syllables that include the low open vowel /a/ are heavy syllables and
generally attract stress.
The phoneme /o/ may have two phonetic realizations, [o] and [ʊ]. The close-mid back
[o] is minimally rounded. The near-back unrounded [ʊ] is observed when the stressed
vowel is followed by [ʧ], [tʲ], [ŋ], [ɳ], and [β̞]. The vowel [ʊ] is often pronounced with
spread lips as in English ‘good’. For example, mapocha ‘papaya’ is realized as [mapʊʧa],
nomotya ‘my belly as [nomʊʧa], paasonki ‘thank you’ as [pa:sʊŋki], onkiro‘rat’ as
[ʊŋkiro], kovitsi [kʊß̞itsʰe] ‘pot’, etc.
The near-minimal pairs representing contrast in analogous environments are given in
(2.3).
2.3
oka ‘this’
aka ‘here’
hiri ‘here it is’
ari ‘that’s alright’
ero ‘negative irrealis’
iro ‘but, nevertheless’
50
2.1.2 Long vowels and diphthongs
Vowel length is distinctive phonemically but the number of words with long vowels is
not high. In addition, vowels can be lengthened due to stress or sometimes in words
borrowed from Spanish, e.g. caro > kaaro ‘car’, camisa > kaamisa ‘shirt’, etc. Examples
with (near) minimal pairs representing contrastive short and long vowels are given in
(2.4).
2.4
piri [pɪɾɪ] ‘your father’
sari [saɾɪ] ‘parrot’
ora [oɾa] ‘that’
entyo [entʲo] ‘woman’s sister’
piiri [pɪ:ɾɪ] ‘bat’
saari [sa:ɾɪ] ‘opossum’
oori [o:ɾɪ] ‘bird species’
eentsi [e:nʦʰɪ̥ ] ‘child’
Vowel length does not have a grammatical function. It can be used as a discourse
strategy to emphasize the extraordinary nature of the event. The word-final vowel is also
lengthened when a vocative form of the kinship term or name is used. Examples
illustrating the lengthening of the word-final vowel in the exclamation clitic =ve and the
kinship term noshiNto ‘my daughter’ are given in (2.5-2.6).
2.5
2.6
Nonyatziro anatarove!
no=ny-atz-i=ro
1SG.A=see-PROG-REAL=3n.m.O
‘I saw the bi-i-i-g one!’
aNtaro=ve
big=EXCL
Noshinto, tzimatsi pinintani?
no=shiNto
tzimatsi pi=niNt-a-ni
1SG.poss=daughter EXIST 2A=want-REAL-AUG
‘Daughter, do you have a fiancé?’
There are two different types of vowel sequences that can be interpreted as
diphthongs: (i) a true phonological diphthong, which is defined as a movement from one
51
vowel to another within a syllable, neither of which dominates (Ladefoged 2001:28;
Clark, Yallop, & Fletcher 2007:66); and (ii) two juxtaposed vowels that belong to
different morphemes. Each of these types is addressed in detail below.
TYPE
(i). There are two types of phonemic diphthongs: centring diphthongs /ia/ and /ea/
and closing diphthongs /ai/ and /oi/, as shown in Table 3. Centring diphthongs are
produced with tongue movement from a peripheral to a central position while closing
diphthongs are produced with tongue movement from a mid or low to a high position
(Clark, Yallop, & Fletcher 2007:35).
TABLE 3. Diphthongs
High
Mid
Low
Front
ia
ea
Central
Back
oi
ai
The marginally attested centring diphthong /ea/ is found syllable-internally in a handful
of words; for example, meanto [mean.to] ‘paradise bird’. The diphthong /ia/, exemplified
by the productive irrealis status suffix -ia, is found word-medially or word-finally, as
seen in (2.7-2.8).
2.7
An.tyoon.kan.tya.ri.
a=N-tyooNk-aNt-ia=ri
1PL.S=IRR-bathe.in.steam-APPL.REAS-IRR=REL
‘In order for us to do steam-bathing.’
2.8
Non.che.kya.
no=N-chek-ia
1SG.S=IRR-cut-IRR
‘I’ll cut myself.’
52
The minimal pairs for /ai/ and /oi/ are pai ‘to call’ vs. poi ‘tree sp. (ashipa)’. There
are no restrictions on the distribution of the closing diphthongs on the syllable or the
word level, as shown in (2.8-2.9).
2.8
no.kai.ma.vai.tzi ‘I kept groaning’
ki.tai.ri.ki ‘peccary’
ai.sa.tzi ‘also’
no.sai.ki ‘I was in it’
iN.pa.kai ‘he will give us’
2.9
in.tsoN.poi ‘inside’
i.tsa.noi.ri.ki ‘his string of beads’
tsoi.ro.ki ‘snail species’
oi.rik ‘to tie’
In the closing diphthongs, the first element is the most prominent while the final
element can be realized as /j/ in fast speech, e.g. paita [paʲta] ‘WH-word’, intsoNpoi
[ɪntsompoʲ] ‘inside’. In fact, the last vowel quality can be so brief and transitory that it is
difficult to identify.
The vowel sequence /ua/, found in some words, e.g. okaNta [o.kuaN.ta] ‘in the
meantime’, Katsiriki [kua.tsʰɨ.ɾɪ.kɪ] ‘a river’s name’, te okaNtyari [te.o.kuan.tʲa.ɾɪ] ‘it
wasn’t so’, may be more economically postulated as a labialized phonetic realization
[kʷ] of the plosive phoneme /k/, which is found when it is followed by the low central
vowel /a/. There are two arguments in favor of this approach. The lip rounding during
the articulation of the velar consonant, conditioned by the phoneme /a/, is found in a few
words only. Additionally, the labialized release of the velar plosive is attested in the
speech of a handful of speakers from an easterly Ashéninka Perené community.
53
TYPE
(ii). This type of vowel sequences refers to two juxtaposed vowels which belong to
two different morphemes such as /ei/, /eo/, /ae/, /ea/. These vowel sequences are not
found among the phonemic diphthongs except for the marginally attested /ea/. Many
examples of vowel clusters on the morpheme boundaries involve the negative particle te,
which is an independent grammatical word, known to form a phonological word with the
vowel-initial verb. Phonetically, these unstressed vowel sequences are often realized in
fast speech as one fused vowel. The morphemic boundaries of the negative particle and
the verb are indicated by the hyphen, as shown in (2.10-2.12).
2.10
/ei/
Te iyote.
te͜-i=yo-t-e
NEG.REAL 3m.S=know-EP-IRR
‘He didn’t know.’
2.11
/eo/
Te oninte onkipatsite.
te-o=niNt-e
o=N-kipatsi-t-e
NEG.REAL 3n.m.S=want-IRR 3n.m.S=IRR-get.soiled-EP-IRR
‘She didn’t want me to get dirty.’
2.12
/ea/
Te amitapintyaro irapinteri irichi.
te-amit-apiNt-ia=ro
ir-apiNt-e=ri
irichi
NEG.REAL 1PL.A.be. accustomed-CUST-IRR=3n.m.O drink-CUST-IRR=REL milk
‘We didn’t drink milk in the past.’
An example of /ae/ is provided in (2.13), where the irrealis suffix –e is adjoined to the
a-final verb stem. As a result, the diphthong /ae/ is formed.
2.13
Pinchakotaena.
p=inchakota-e=na
2A=split-IRR=1SG.O
‘Split wood for me.’
54
2.2 Consonants
2.2.1 Phoneme inventory
Table 4 represents 20 consonant phonemes of Ashéninka Perené, given in practical
orthography (see §2.10). The corresponding phonetic values are provided in square
brackets. Here follow some comments on Table 4.
(i) One feature that is immediately noticeable is the lack of phonemic voice
distinction in Ashéninka Perené. Voiced stops, fricatives or affricates are non-existent;
neither are present voiceless liquids or nasals.
(ii) There is an opposition between aspirated and unaspirated voiceless affricates.
(iii) There is a set of voiceless oral stops which contrast with their palatalized
counterparts.
(iv) Palatalization applies to all stops and alveolar phonemes for all manner of
articulation such as stops, affricates, fricatives, liquids, and nasals. Note that sets of
palatalized stops or alveolar fricatives are found in some Arawak and Amazonian
languages (Aikhenvald 1999:76; Guillaume 2004:27).
(v) There is a contrast between nasal consonants /m/, /n/, and /ɲ/, and /N/, a nasal that
is underspecified for place of articulation, addressed in detail in §2.2.6.
Nearly all consonant phonemes occur syllable-initially, except for the underspecified
nasal N, which occurs in a syllable-final position before a plosive. Syllable-final
consonants are restricted to the nasal stop /n/ and the archiphoneme N. Consonants do not
occur word-finally, except for ideophones and unassimilated Spanish loans.
55
TABLE 4. Consonant phonemes and their allophones
Voiceless stops
palatalized
Voiceless affricates, aspirated
unaspirated
Voiceless fricatives
Liquids
Nasals
Semivowels
Archiphoneme
2.2.2
bilabial
p [p, pʲ]
py [pʲ]
velar
k [k, kʲ, kʷ, g]
ky [kʲ]
m [m]
v [w, ß̞]
N
alveolar
t [t, tʰ, d]
ty [tʲ]
ts [ʦʰ]
tz [ʦ]
s [s]
r [ɾ]
n [n, ŋ]
palatal
glottal
ch [ʧ]
sh [ʃ, s]
ry [ɾʲ]
ny [ɲ]
y [j]
h [h, hʲ]
Stops /p/, /t/, /k/ and their palatalized counterparts /pʲ/, /tʲ/, and /kʲ/
The non-palatalized stops in Ashéninka Perené are pronounced similarly to their
equivalents in other languages, e.g. in Spanish. The palatalized stops are realized
similarly to their equivalents in Russian. The distribution and phonetic realizations of the
stops are outlined below.
2.2.2.1 The phonemes /p/ and /pʲ/
The phoneme /p/ is a voiceless, unaspirated bilabial stop, pronounced similarly to
Spanish /p/ in poner ‘put’. It occurs word- and syllable-initially as in potooki ‘tree sp.
(ojé)’, opempe ‘tucan’, intsipa ‘guaba’, shimpoki ‘horsefly’, chompita ‘cockroach’. Its
default realization is [p]. It is in complimentary distribution with [pʲ] attested with i-final
verb roots, e.g. root pi ‘lose’, ‘convert’, at the morpheme boundary when the root is
followed by the non-front vowel [a]. Examples (2.18-2.19) show the mutually exclusive
environments in which the allophones [p] and [pʲ] appear.
56
2.18
Ipiiri.
i=pi-i=ri
3m.A=transform-REAL=3m.O
‘He was converting them.’
2.19
Ipyanaka maniro.
i=pi-an-ak-a
3m.S=transform-DIR-PRF-REAL
‘He transformed into deer.’
The phoneme /pʲ/ is a palatalized voiceless bilabial plosive, whose distribution is
identical with that of its counterpart /p/. The contrast between /p/ and /pʲ/ is demonstrated
in (2.20).
2.20
-tsapaki ‘finger’
tampatsika ‘be vertical’
paryantsi ‘plantain’
tsapyaki ‘on the river bank’
tampyaa ‘wind’
pyarentsi ‘masato [manioc beer]’
2.2.2.2 The phonemes /t/ and /tʲ/
The phoneme /t/ is a voiceless alveolar stop found word- and syllable-initially as in tashi
‘roast’, apite ‘two’, potoNto ‘one having mucus running from his/her nose’, etc. Its
pronunciation is similar to that of its Spanish counterpart in todo ‘all’. It has three
allophones [tʰ],[d], and [t]. In previous studies of Ashéninka Perené, the alveolar
phoneme /t/ was attested to have an allophone [tʰ] (Payne 1989:23). This study has
revealed that the allophone [tʰ] occurs in the speech of older speakers when followed by
the low central vowel /a/, e.g. vathatsa [watʰatsa] ‘meat, flesh’, nothanoirikipaye
[notʰanoiɾikipaje] ‘my adornments’. The phoneme /t/ takes on voicing in the speech of
some speakers when it occurs intervocalically, e.g. shetene ‘day’ is pronounced [ʃεdεne].
In other environments, [t] is the default. There is also an independent phoneme /tʲ/, which
57
has the same distribution as its counterpart /t/. Examples in (2.21) demonstrate a contrast
between /t/ and /tʲ/.
2.21
paamaNto ‘burned wood’
notaapi ‘my back’
=ta ‘subjunctive clausal enclitic’
pantyo ‘duck’
tyaapa ‘chicken’
=tya ‘emphatic clausal enclitic’
2.2.2.3 The phonemes /k/ and /kʲ/
The phoneme /k/ is a voiceless velar unaspirated stop. It is found word- and syllableinitially but can occur word-finally in ideophones, e.g. tsinik ‘dropping sound’, tsiNpak
‘sound of an agonizing human body’. It is pronounced similarly to its Spanish counterpart
in cama ‘bed’. Examples illustrating distribution of the phoneme are the following: kaniri
‘manioc’, kashiri ‘moon’, okaakini ‘near’, mapirotaka ‘she is worse’.
The phoneme /k/ has four allophones, [k], [kʲ], [kʷ], and [g]. The allophone [pʲ] is
attested with i-final verb roots, e.g. ki ‘carry’, load’, at the morpheme boundary when the
root is followed by the non-front vowel [a], exemplified by (2.22-2.23).
2.22
Ikiiri.
i=ki-i=ri
3m.A=carry-REAL=3m.O
‘He was carrying him on his back.’
2.23
Ikyanakina.
i=ki-an-ak-i=na
3m.A=carry-DIR-PRF-REAL=1SG.O
‘He carried me on his back.’
The phoneme /k/ is pronounced by some speakers as [kʷ] when followed by the low
central vowel /a/, e.g. okaNta [okʷanta] ‘in the meantime’. The attested labialization of
58
the velar stop in this environment occurs irregularly even in the speech of the same
speaker.
The voiceless phoneme /k/ is realized by some speakers as its voiced counterpart [g]
when found after the nasal N and followed by a vowel, or when it is found between
vowels, as seen in (2.24) below.
2.24
nasaNkaniki
kityoNkari
opaNkita
katoNkosatzi
nairikaka
nochekakiro
naaka
[nasaŋganɪkɪ]
[kɪtʲoŋgaɾɪ]
[opaŋgɪta]
[katoŋgosatsɨ]
[naɪɾɪgaka]
[noʧegakɪɾo]
[na:ga]
‘in my heart’
‘red’
‘it is planted’
‘people from upriver’
‘I held on (to the tree)’
‘I cut it’
‘I’
The [k]~[g] alternation is generally observed in casual speech and is speaker-dependent.
The contrast between the phoneme /k/ and its palatalized counterpart /kʲ/ (which has the
same distribution as /k/) is evident in (2.25).
2.25
kyaarontsi ‘winter’
kyaaryo ‘it’s true!’
=kya ‘emphatic enclitic’
inkaari ‘rain’
kaari ‘negative copula’
=ka ‘interrogative clausal enclitic’
2.2.3 Affricates ts /tsʰ/, tz /ts/, and ch /ʧ/
There are two alveolar voiceless phonemic affricates ts /tsʰ/ and tz /ts/, the latter marked
by the absence of aspiration. The unaspirated affricate /ts/ is pronounced similarly to the
/ts/ sound in English pizza. Note that the unaspirated tz /ts/ is predominantly realized in
the westerly speech variety. Some speakers of the easterly speech variety do not
distinguish between the two, using only the aspirated affricate ts /tsʰ/, pronounced
59
similarly to the Armenian aspirated affricate /tsʰ/ in bayts ‘but’. The affricates are found
syllable- and word-initially and word-medially. The contrast between ts and tz is
illustrated in (2.26).
2.26
tzina ‘palm sp.’(camonilla)
tzivana ‘pineapple’
tziNkotsi ‘tail’
tsinani ‘girl’
tsiva ‘tree sp.’ (copal; Bursera odorata)
tsiNt ‘to urinate’
There is one palato-alveolar voiceless affricate ch /ʧ/, pronounced similarly to the
word-initial affricate in English ‘chip’. Akin to the loss of the distinction between the
alveolar affricates ts and tz (they are both pronounced with some degree of aspiration),
there seems to be a similar blending process of merger underway involving the
palatalized alveolar stop /tʲ/ and the palato-alveolar affricate /ʧ/. The two are converging
perhaps due to the natural tendency in the articulation of a palatalized stop to produce
some degree of air turbulence and friction (i.e. affrication) (Clark, Yallop, & Gletcher
2007:63-4). Contrasts can be demonstrated for /ʧ/, /tsʰ/ and /tʲ/ with the following
examples in (2.27).
2.27
/ʧ/-/tsʰ/ chomeNta ‘fish sp.’
/ʧ/-/tʲ/
kachori ‘salted’
/tsʰ/-/tʲ/ ipitsoki ‘he was turning around’
tsomoNte ‘a bulge’
katyori ‘acidic, sour’
ipityoki ‘his little thing’
2.2.4 Alternative analyses of palatalized segments and affricates
Alternative analyses of the palatalized consonants and affricates can be suggested, but it
will be shown below that they turn out to complicate the explanation of the data. One
such alternative is to abandon the analysis of the palatalized consonants pʲ, tʲ, kʲ, rʲ, ɲ as a
60
distinct series of unitary phonemes and instead posit them as consonant clusters made up
of consonant-consonant segments Cj, that is pj, kj, rj, nj. Similarly, we can interpret
affricates tsʰ, ts, and ʧ as consonant clusters CC consisting of separate phonemes, /t/, /s/
and /t/, /ʃ/.
In favor of such alternative analyses is that each consonant is found outside the
consonant sequences, except for the aspirated affricate tsʰ. Note that the language lacks
aspirated consonants, which renders problematic an analysis of this affricate as a
consonant cluster. The alternative analysis is flawed for a number of other reasons as
well. There are no native consonant sequences such as sp, st, pl in the language. If we
posit a CC interpretation for these segments, it will be difficult to explain why these
consonant clusters are allowed in onsets while others are not. We will also need to posit a
more complex syllable structure CCVN to account for these phenomena. However, if we
allow the original interpretation of the unitary nature of the palatalized consonants and
affricates, we get a simple, consistent, elegant system in which a single feature accounts
for the data. We get a series of stops, affricates, fricatives, nasals, liquids and their
palatalized counterparts /p/, /pʲ/; /t/, /tʲ/; /k/, /kʲ/; /tsʰ/, /ʧ/; /tz/, /tʲ/; /s/, /ʃ/; /n/, /ɲ/, and /ɾ/,
/ɾʲ/. Similar series of palatalized segments, which appear to be uncommon among
Amazonian languages, are found in other Kampan languages (Michael 2008:222).
2.2.5 Fricatives s /s/, sh /ʃ/, and h /h/
There are three voiceless phonemic fricatives, two sibilants /s/ and /ʃ/, and a glottal
fricative /h/. The fricatives occur word and syllable-initially and word-medially as in
samani ‘paca’, imposaiteri ‘they will hit him’, shima ‘fish’, inchashi ‘tree leaf’, hevari
61
‘chief’, kahai ‘coffee plant’. The phoneme /s/ is an alveolar voiceless fricative,
pronounced similarly to [s] in Spanish sabor ‘taste’.
There is an independent alveo-palatal phoneme sh /ʃ/, found in the same
environments as /s/. It is pronounced similarly to [ʃ] in Spanish ishpingo ‘cedar
mahogany tree’. In general, data from speakers of Kampan languages is suggestive of the
regular palatalization process when sibilants that follow high front vowels are palatalized;
that is s→ʃ before i, but in other environments /s/ and /ʃ/ contrast (Payne 1983: 102-3).
Payne notes that Alto Perené speakers articulate /s/ in all positions, including before i
(1989: 23). In our data, this observation holds only for some older speakers who say
[pasɪ̥ ne] ‘other’, [sɪ̥ ma] ‘fish’, etc. Older speakers also pronounce the sibilant phoneme as
[sʲ] when it followed by the sequence /iy/, e.g. ‘Take a picture of it’ is pronounced
[pɪsʲijakantapaɪnteɾo] by older speakers while younger speakers say it
[pɪʃijakantapaɪnteɾo].
In addition, the sibilant phoneme can be realized as [h], e.g. pishetopaye [pɪhetopaje]
‘your intestines’ (it is not clear whether this phoneme is underlyingly /s/ or /ʃ/ for the
language speakers). With other words, /ʃ/ is generally realized as [ʃ], e.g. noshirontahyari
[noʃɪ̥ ɾontohʲaɾɪ] ‘I will smile at it’; oshoNkapakiro [oʃoŋkapakɪɾo] ‘she went around it’.
Examples in (2.28) demonstrate the contrast between /s/ and /ʃ/.
2.28
samani ‘paca (rodent sp.)’
kasaNto ‘bird species’
saNkina ‘to write’
shametzi ‘pal; namesake’
ashaniNka ‘our fellowman’
shiNki ‘corn’
62
The glottal voiceless fricative /h/ is pronounced like [h] in Spanish jaca ‘pony’, e.g.
paahe pashini ‘you’ll take another [woman]’. It is pronounced as [hʲ] when followed by
the irrealis suffix –ia, e.g. ikinashihetanahya ‘they walked in different directions’.
2.2.6 Nasals m /m/, n /n/, ny /ɲ/, and the archiphoneme /N/
There are three phonemic voiced nasal stops in Ashéninka Perené, one bilabial stop /m/,
one alveolar stop /n/, and a palatal stop /ɲ/. The nasal stops are found word- and syllableinitially, syllable-finally, and word-medially.
The bilabial stop /m/ is generally realized as [m] as in mava ‘three’, =mi 2SG
personal enclitic, yomit ‘ teach’, nomotya ‘my belly’, samaNpo ‘ashes’, aritaima
‘maybe’. The alveolar nasal /n/ is realized as [n] as in nosaro ‘grandmother’, incha
‘plant’, pashini ‘other’, nimot ‘to like’, anoNk ‘to swell’, pairani ‘long ago’.
The palatal nasal /ɲ/ is realized as [ɲ], e.g. nyaantsi ‘word’, ironyaaka ‘now’, oponya
‘then’.
The following list in (2.29-2.30) shows contrast between /m/ and /n/, and /n/ and /ɲ/.
2.29
/n/ vs. /m/
kanai ‘Long John tree’
kin ‘to pass’
kamaantsi ‘phlegm’
kim ‘to hear, listen’
2.30
/n/ vs. /ɲ/
naakero ‘I’ll bring it’
nyaakiro ‘I’ve seen it’
kona ‘to fish with barbasco’ konya ‘to appear’
Syllable-finally, the nasal segmental unit assimilates homorganically to a following
plosive, as demonstrated in (2.31).
63
2.31
onaNpi [o.nam.pi] ‘her village’
kasaNto [ka.san.to] ‘bird sp.’
katoNko [ka.toŋ.ko] ‘upriver’
mapi [ma.pɪ] ‘stone’
onatoto [o.na.to.to] ‘cousin’
nako [na.ko]‘my arm’
This situation when the underlying nasal has the surface realization [m, n, ŋ] in certain
environments suggests that the nasal is unmarked (or unspecified) for the place of
articulation. This phoneme is posted as a separate archiphoneme /N/ since it is distinct
from other nasals (it only occurs following a vowel and assimilates to a following
obstruent). Further evidence comes from the spelling habits of language consultants, who
consistently write all syllable-final nasals either with m or n unless they are taught to
make a distinction between the two. This circumstance points at the syllable-final nasals,
though varied in phonetic form, as being not derived from the syllable-initial nasals;
rather, they constitute an independent, unspecified for place of articulation segment. A
detailed discussion of the archiphoneme /N/ in Ashéninka and other Kampan languages is
found in Burquest (2006:67-8), Michael (2008:223-4), and Payne (1981: 62-3).
2.2.7 Liquids r /ɾ/ and ry /ɾʲ/
There are two liquids in Ashéninka Perené, namely a voiced alveolar flap /ɾ/ and its
palatalized counterpart /ɾʲ/, a typical situation for an Arawak and Amazonian language
which generally have one or two liquids (Aikhenvald 1999:77).
The alveolar flap /ɾ/ is pronounced close to the Spanish liquid in pero ‘but’. The
phoneme can occur syllable-initially and word-medially but never word-initially as in
iriro ‘he’, irika ‘this (masculine), ari ‘positive polarity verb’, naari ‘I’, aparoni ‘one’,
64
airo ‘not’, etc. This restriction on the word-initial placement of /ɾ/ applies to Spanish
loans as well, e.g. irichi > leche ‘milk’, iribiro > libro ‘book’.
The phoneme /ɾʲ/ resembles the Russian palatalized liquid in /hvorʲ/ ‘ailment’. Its
distribution is identical to that of its non-palatalized counterpart /ɾ/ as in piryoki ‘coconut
palm grub’, paryantsi ‘plantain’. The contrast between the /ɾ/ and /ɾʲ/ is evident in (2.32).
2.32
piryoki ‘grub’
paryantsi ‘plantain’
pirota ‘ball’
parari ‘otter’
2.2.8 Semivowels v /w/ and y /j/
There are two glide phonemes /w/ and /j/ in Asheninka Perené. The phoneme /w/ is a
bilabial approximant. Before non-front vowels, it is pronounced as [w], with minimal lip
rounding. It occurs syllable- and word-initially or -medially as in varipa ‘chicken’ and
pava ‘god’. Before front vowels /i/ or /e/, /w/ is realized as a bilabial voiced approximant
[ß̞], resembling the medial sound in Spanish ave [aß̞e] ‘bird’. It is pronounced with spread
lips, as shown in Figure 10.
FIGURE 10. Articulation of the bilabial approximant in ivito ‘his canoe’
Occasionally, the bilabial approximant [ß̞] is articulated as a labio-dental approximant
[ʋ]. Examples are given in (2.33).
65
2.33
kavana
avavatzi
ovakera
[kawana] ‘tree sp.’
[awaʋatsi] ‘we are eating’
[owakeɾa] ‘recently’
tzivi
vatsatsi
kovitsi
[tsɨß̞ɪ]
[watsʰatsʰɪ̥ ]
[koß̞itsʰe]
‘salt’
‘meat’
‘pot’
In Figures 11-13, we can compare spectra of pava ‘god’, novori ‘my muscle’, and
ivito ‘his canoe’ respectively which contain an intervocalic bilabial approximant. Figures
11-13 show that the first and second formants of [w] are low in comparison with the
higher second formant of [β̞].
FIGURE 11. Phonetic realization of the bilabial approximant in pava ‘god’
p
a
w
a
FIGURE 12. Phonetic realization of the bilabial approximant in novori ‘my leg’
n
o
w
o
ɾ
i
66
FIGURE 13. Phonetic realization of the phoneme /w/ in ivito ‘his canoe’
i
ß̞
i
t
o
The bilabial voiced unrounded approximant may be alternatively treated as an
underlyingly voiced bilabial fricative fricative /β/, as seen in Romani (2003) who posits
the bilabial approximant as a voiced bilabial fricative in the closely related Kampan
language Ashéninka Pajonal. However, this analysis argues against this approach based
on the following argumentation. The consonant inventory of Ashéninka Perené does not
have voice distinction in stops, fricatives and affricates, which are all voiceless, whereas
its bilabial and palatal phonemic approximants /w/ and /j/ are voiced. Phonemic voice
distinctions in bilabial stops are mainly found in non-Kampan Arawak languages and in
some Kampan Arawak such as Nanti (Michael 2008) and possibly, in Nomatsiguenga
(Aikhenvald 1999:76). As far as bilabial fricatives are concerned, they are rare, though a
voiced bilabial fricative β is found in the adjacent non-Kampan variety Amuesha
(Aikhenvald 1999:76). In contrast, other Amazonian Arawak languages have bilabial
approximants, e.g. Ashéninka Apurucayali and Cavineña have a phonemic bilabial
approximant /w/ with the allophones [w] and [ß̞] (Black 1991:185; Burquest 2006:92-4,
159; Chitoran & Nevins 2008:1900; Guillaume 2004:29-30; Payne, Payne, & Santos
1982:77). Evidence in support of the proposal that the underlying phoneme is /w/ and not
67
a bilabial stop or fricative is that there is another approximant in the language, the palatal
glide /j/ which behaves similarly to /w/ and can be considered its symmetrical
counterpart. In addition to sharing the feature [+voice], both glides have a similar
distribution, and show consonantal behavior when they fill the onset position in syllable
templates. The palatal glide y /j/ occurs word- and syllable-initially and word-medially as
in yaya ‘woman’s brother’, yaarato ‘bee’, yentsiri ‘tree sp.’, ayoNpari ‘friend’, noyeNpita
‘my ear’, -paye ‘nominal plural’. There are no other voiced stops or fricatives in the
language that would share symmetry with the bilabial voiced approximant.
Another alternative interpretation of glides /w/ and /j/ as vowels /u/ and /i/
respectively can be suggested, although it appears to present certain difficulties for the
explanation of the data. If we favor the vowel sequence analysis such as /ua/, /ue/, /uo/,
/ui/, we make more complex the inventory of diphthongs. We also violate the
requirement for onsets in medial syllables since Ashéninka Perené disallows onsetless
syllables word-internally e.g. a.va.va.kʲa ‘we’ll eat’ * a.ua.ua.kʲa. Another requirement,
of no more than two adjacent vowel segments in vowel sequences, will be violated as
well. In addition, we will have to ignore the allophonic variation of the phoneme /w/ if
we were to allow the vowel sequence interpretation.
The attraction of the alternative analysis of the glide /j/ as a high front vowel /i/ lies in
the fact that then we get a resultant drastic simplification of the phonemic inventory by
disposing of the series of palatalized stops, fricatives, nasals, liquids (while only keeping
tsʰ, ts, and ʧ). Note that literate Ashéninka Perené speakers write palatalized consonants
as Ci sequences, e.g. piarentsi ‘plantain’. This analysis was also advocated by the SIL
linguist Judith Payne (1990). Nevertheless, a convincing piece of evidence comes from
68
acoustic analysis which supports an interpretation of palatalized consonants as unitary
segments, rather than a consonant-vowel sequence Ci. In Figure 14, which is the
spectrogram of palatalized alveolar stop in the word-medial position, the red line marks
the consonantal release from the palatal position when the tongue body assumes an i-like
position. The spectrogram of this transition is distinct from spectra with a sequence of Ci
or Cj (see Ladefoged & Maddieson 1996:364-5).
FIGURE 14. The palatalized stop /tʲ/ in novityero ‘I’ll fell it (the tree) ’
n
o
ß̞
ɪ
tʲ
e
ɾ
o
2.3 Sounds in Spanish loanwords
Ashéninka Perené speakers have borrowed a lot of words from Spanish, a deplorable fact
in the eyes of many native speakers. All speakers I worked with are highly bilingual and
do a great deal of intersentential code-switching and intrasentential code-mixing while
talking among each other. Most of the Spanish loans are nativized in Ashéninka Perené,
i.e. fully adapted to the structure of the recipient language. This process of nativization
involves adaptation of lexical material to the morphological, syntactic and phonological
patterns of the language (Thomason 2001:134). Examples of nativized Spanish loans are
69
listed in (2.34).The first column lists the Spanish lexemes and the second shows the
corresponding Ashéninka Perené loans.
2.34
a.bol.sa ‘bag’
b. pa.pel ‘paper’
c. le.che ‘milk’
d. pas.ti.lla ‘pill’
e. pe.lo.ta ‘ball’
f. hos.pi.tal ‘hospital’
g. re.loj ‘watch’
h. a.rroz ‘rice’
bo.ri.sa ‘bag’
pa.pe.ri ‘paper’
i.ri.chi ‘milk’
pa.shi.tya ‘pill’
pi.ro.ta ‘ball’
o.si.pi.ta.ri
no.re.ro.te ‘my watch’
a.ro.so ‘rice’
The phonological structure of the borrowed material is assimilated by the speakers
based on the Ashéninka Perené phoneme inventory (see Table 4), canonical syllable
shape (see §2.4), and stress assignment patterns (see §2.5). As (2.34) shows, the Spanish
liquid /l/ is replaced with the Ashéninka Perené flap /ɾ/ in all loans. If the Spanish liquid
is word-initial, it is replaced with the Ashéninka flap /ɾ/ and follows a vowel, as in
(2.34c). Word-final consonants are disallowed in Ashéninka Perené, which explains the
addition of vowels word-finally in nativized loans in (2.34 b, e-i). Consonant clusters like
sp, st are not permitted either, resulting in the replacement of the consonant cluster st
with a simplex fricative ʃ in (2.34d). In terms of stress, Asheninka Perené nouns typically
assign second-syllable stress while the stress in the borrowed Spanish words may fall on
the last syllable as in (2.34d).
There is a sizable number of unassimilated loans in Ashéninka Perené, e.g. escuela
‘school’, estado ‘state’, contrato ‘contract’, particular ‘private’ which violate native
phonotactics by allowing consonant clusters and/or by ending in a consonant. Numerals
borrowed from Spanish do not comply with phonotactics either. They are borrowed
70
without any change in their pronunciation, e.g. tres mil ‘three thousand’. The
unassimilated loans also preserve their Spanish orthography when written by literate
Ashéninka Perené speakers.
2.4
Phonotactics
When defined narrowly, phonotactics refers to a distributional statement describing how
segments are distributed within syllables, syllables withing words, etc. In a broader sense,
phonotactics is known as a general description of sequences and combinations.
Phonotactic constraints express limitations on the free combinations of units (Clark,
Yallop, & Fletcher 2007:69). Constraints on the distribution of phonemes are found on all
three levels of segmental phonology. The first level, the level of segment adjacency,
shows limitations on some segments co-occurring contiguous to others. Vowel sequences
of more than two segments are not allowed. Some vowel sequences do not occur, e.g. ie,
oe. Consonant clusters are disallowed. A few consonant-vowel combinations never cooccur: ty /tʲ/ and ry /rʲ/ never occur before /i/.
The gaps in the distribution of phonemes in consonant-vowel sequences are very few.
Arguably, these gaps can be attributed to articulatory difficulty with a sequence. For
example, /tʲi/ and /ɾʲi/ are probably difficult to articulate for the speaker. The alveopalatal
segments are complex transitional articulations involving the constriction of the basic
articulation and raising the tip and blade of the tongue to a high front position as part of
the transition to the next segment (Clark, Yallop, & Fletcher 2007:64). When the next
element is a high front vowel /i/, the palatal approximation of the tongue tip and blade is
71
likely to be difficult in speech production; hence the constraint on these alveopalatal
consonant-high front vowel sequences.
Ashéninka Perené shows certain restrictions on the next level, the level of distribution
of the segments within a syllable. The following limitations, found in other Kampan
languages, e.g. in Nanti (Michael 2008: 239-40), are listed below.
(i) No complex onsets or codas are permitted; when a formative beginning with a
consonant attaches to a unit ending in a consonant, the epenthetic segment a is inserted.
(ii) Although consonantal onsets are allowed, consonantal codas are disallowed, except
for nasals which can occur in the coda; in addition, syllable-final consonants are found in
ideophones and unassimilated Spanish loans.
(iii) Onsetless syllables are restricted to the word-initial position.
(iv) No hiatus is allowed word-internally; when a formative beginning with a vowel
attaches to a unit ending in a vowel, the epenthetic segment t/tz is inserted.
At the word level, consonants are never found word-finally in native Ashéninka
Perené words, except for a few onomatopoeic forms. The flap ɾ (or its palatalized
counterpart ɾʲ) never occurs word-initially.
The preference stated above for avoidance of morpheme-internal clusters is illustrated
by assimilated loans from Spanish such as borisa derived from bolsa ‘bag’, osipitari
created from hospital ‘hospital’. Yet consonant clusters are accepted in a number of
unassimilated loans as in estado ‘state’, escuela ‘school’, amburancia ‘ambulance’. The
constraint on word-final consonants is violated in other borrowed Spanish words such as
mision ‘misión’, pastor ‘pastor’, diez ‘ten’, mil ‘thousand’.
72
2.4.1
Syllable structure
The basic syllable structure is CV, although the minimal syllable shape is V. Other
permitted syllable shapes are VN, CVN, CVVN, CVV, and VV. Only the vowel nucleus
V is obligatory, although this syllable shape is not common; it is found solely wordinitially. The syllable shape does not allow for syllables larger than CVVN. Examples of
Ashéninka Perené syllable shapes are provided in (2.35). Syllables in (2.35) are separated
by a dot.
2.35
CVV.CV.CV.CV
V.CV
VV.CV
VN.CV
VN.CVN.CVV
CVVN.CVVN.CV
poi.ni.ro.tsa [poɪnɪɾotsʰa]
a.ri
[aɾɪ]
ai.sa.tzi
[aɪsatsɨ]
in.cha
[ɪnʧa]
in.tsoN.poi
[ɪntsʰompoi]
tyooN.kaan.tsi [tʲoŋka:ntsʰɨ]
‘yacón (tree sp.)’
‘positive polarity verb’
‘also’
‘plant’
‘inside’
‘to bathe in steam’
Since Ashéninka Perené strongly prefers word-internal syllables to have onsets, it uses
epenthetic segments a (inserted between the consonants at the syllable/morpheme
boundary) or t/tz 10(inserted between vowels) to produce well-formed syllables as in
(2.36-2.37).
2.36
Y=ook-ai-t-a=na.
yoo.kai.ta.na
3m.A=leave-IMP.P-EP-REAL=1SG.O
‘They left me.’
2.37
No=kaim-a-vai-tz-i.
no.kai.ma.vai.tzi
1SG.S=call-EP-DUR-EP-REAL
‘I kept groaning.’
10
The epenthetic segment /t/ has two allophones, [t] and [ts]. The latter is conditioned by the high front
vowel /i/ when it follows the epenthetic segment.
73
However, this requirement can be overridden when vowel-initial V or VC suffixes such
as –a,-ak, -ah, -ite are adjoined to the vowel-final root, as seen in (2.38-2.39).
2.38
Naakiro.
n=a-ak-i=ro
1SG.A=take-PRF-REAL=3n.m.O
‘I took it’.
2.39
Haitetaki.
ha-ite-t-ak-i
go-quickly-EP-PRF-REAL
‘She went quickly.’
Comparative studies of Kampan languages provide evidence that the root a ‘take’ in
Proto-Kampan used to have a consonantal coda; the suffix -ite ‘quickly’ most likely had a
consonantal onset. In spite of the segmental reduction the roots and suffixes have
presumably undergone, they do not participate in the epenthetic processes.
The phonotactics of the language disallows consonant clusters word-internally but
they occur during the surface realization of some words due to the elision of the
unstressed weak vowel /i/ when it occurs between sibilants, affricates, and voiceless
plosives as in antyashipari [antʲaʃpa.ɾi] ‘elder’, oshipitari [oʃpitaɾi] ‘hospital’, tsika
[tsʰka] ‘where’.
2.4.2 The minimal word
The minimal word in Ashéninka Perené must have two syllables at minimum. Another
requirement for a minimal word is an independent stress. The syllable shapes of the
minimal prosodic words allowed in the language are given in (2.40).
74
2.40
a. CV.CV
b. V.CV
c. V.CVV
d. CVV.CV
̍hi.ri
̍a.ri
i.̍roo
̍naa.ri
‘here he is’
‘positive polarity verb’
‘she’
‘I also’
The only exceptions appear to be the monosyllabic CV negative and affirmative
particles te and he containing a short vowel /e/. The particles can be used independently
as responses to the polar questions. The negative particle te is stressed in (2.41).
2.41
-Notomi, omapiro piposakiri paniri?
-Te, apaa.
-Son, is it true that you hit your cousin?
-No, Dad.
There is also a monosyllabic stressed ideophone poo with the syllable shape of CVV
which imitates the hitting sound. An example is provided in (2.42).
2.42
Oposakina noitoki poo!
o=pos-ak-i=na
no=ito-ki
poo!
3n.m.A=hit-PRF-REAL=1SG.O 1SG.poss=head-LOC IDEO
‘It hit my head, bam!’
The negative and affirmative particles te and he and the ideophone poo are examples
of independent grammatical words (see §3.1 for discussion of grammatical vs.
phonological words) which tend to attach to the adjoining word form.
Nouns are generally disyllabic or trisyllabic, e.g. ka.̍hai ‘coffee plant’, ̍mi.chi
‘cat’, k̍ o.ta ‘piece’, sha.̍ve.ta ‘butterfly’, che.̍vo.tsi ‘branch’. Other lexical categories
such as pronouns, adjectives, adverbs, demonstratives, particles are at least disyllabic as
in n̍ aa.ka ‘I’, ̍kaa.ri ‘negative existential’, ro.̍ka ‘that (non-masculine), ̍ai.ro ‘not
(irrealis)’, ̍pai.ra.ni ‘long ago’. Verbs must have at least two syllables because of the
obligatory subject and reality status marking as in pi.ya! ‘Eat!’ which consists of pi= 2SG
75
subject proclitic and the irrealis suffix –ia, fused with the verb root v ‘eat’. Grammatical
morphemes, e.g. aspect, reality status markers are often monosyllabic, as well as
applicatives, classifiers, and clitics but they never form an independent word.
2.4.3
Stress
Stress is not phonemic in Ashéninka Perené. The prominence of a stressed syllable is
normally represented by its loudness, higher pitch, and longer duration. Stress patterns in
Ashéninka Perené are varied. In independent bisyllabic words, which are not cliticized or
bound to other words, stress may occur on the either syllable, as seen in (2.43).
2.43
ni.ˈhaa
'niN.ka
a.ˈka
‘water’
‘who’
‘here’
i.ˈroo
taN.ˈpyaa
ˈnaa.ka
‘she’
‘wind’
‘I’
Stress assignment in (2.43) is sensitive to syllable weight. Syllable weight, ranked
from the highest to the lowest CVVN> CVV> CVN> CV> V, tends to determine stress
assignment in independent words. As seen in (2.41), syllables with branching nuclei (a
long vowel or a diphthong), and/or a coda ((C)VV, CVN, (C)VVN) are likely to attract
stress, but there are many exceptions, e.g. ˈko.ta ‘sole’, choo.ˈki ‘sister’.
Stress also depends on the sonority of nuclei. Based on the sonority scale of
Ashéninka vowels, arranged from the strongest to the weakest, a> e, o> i, the stress is
likely to be assigned to the syllable with the stronger sonority as in (2.44).
2.44
shi.ˈma
tzi.ˈna
shi.̍ na
‘fish’
‘palm sp.’
‘kapok tree’
i.ˈsha
mi.ˈsha
tsi.ˈka
‘grandmother’
‘naked tree (capirona)’
‘where’
76
In independent trisyllabic words, the stress normally falls on the penultimate syllable,
as seen in (2.45). Syllable boundaries are shown by a period.
2.45
o.ˈche.vo
ka.ˈme.tsa
pa.ˈshi.ni
‘its branch’
‘good’
‘other’
ko.ˈshi.ri
paa.ˈma.ri
tzi.ˈva.na
‘monkey’
‘fire’
‘pineapple’
The penultimate stress pattern can be violated due to the syllable weight constraint as
in shin.tsi.ˈpaa ‘raft’ where the first two light syllables are outranked by the last syllable
with the branching nucleus. In assimilated Spanish loans, the stress tends to follow the
penultimate stress pattern, e.g. so.ˈra.ro ‘police [Spanish soldado]’, bo.ˈri.sa ‘bag’
(Spanish bolsa). In polysyllabic loans, stress assignment follows the same rule as in
o.shi.pi.ˈta.ri ‘hospital’. In general, in bi- and trisyllabic words, the primary stress is
assigned to the ultimate syllable if it is heavy, and otherwise to the penultimate syllable.
In polymorphemic words, the general principle of avoidance of stress on the outer
edges of the word is observed. As discussed in detail in §3.3.2, the noun template slots
may be filled with multiple formatives on both sides of the root. The verb predicate
structure may also be composed of multiple morphemes. Peripheral formatives such as
person, plural, possessive markers, etc. are never stressed in polymorphemic
phonological words. Given the theoretical assumption that certain formatives are ignored
in building metrical structure, we consider such material extrametrical, when not used for
stress assignment (Burquest 2006:282). Examples of stress patterning in polysyllabic
words are shown in (2.46). Extrametrical segments are in angle brackets.
77
2.46
̩
ya.shi.ˈro.tsa.<te >
i.va.ˈryan.tsi.<te.pa.ye>
no.̩sai.ka.ˈhe.tzi.<ka>
̩a.ka.̩ra.he.ˈtai.ye.<ni>
‘his ax’
‘his plantains’
‘we are here’
‘we all’
Generally, the rightmost left-headed foot serves as head of a polysyllabic word
whereas stress feet to the left take secondary stress (the secondary stress is not very
salient). A phonological definition of a foot refers to a primarily disyllabic rhythmic unit
with one most prominent, stressed syllable serving as its head. If the rhythmically strong
syllable is followed by a rhythmically weak syllable, the foot is left-headed. While the
prominence of the primary stress is associated with the penultimate syllable, the
secondary stress is associated with every other syllable to the left of the primary stress.
Overall, the complexity of rhythmic stress patterns in Ashéninka Perené is far from being
well understood at this point. For more details on stress patterns in Kampan (Ashéninka
and Nanti), see Payne 1990 and Crowhurst & Michael 2005.
2.5 Intonation
There is no special intonation at the end of declarative clauses. Generally, falling
intonation signals the end of a clause. Greetings are pronounced with a falling intonation,
as in kitaiteri ‘Good morning’, sheteneri ‘Good afternoon’, tsiteneri ‘Good evening’,
hatanara ‘So long [I am going]’. Some greetings have a rising intonation contour, which
marks the stressed final syllables as seen in (2.47). The stressed syllables are given in
bold.
2.47
-Naakave! ‘Hi! [I am]!’
-Aviro!
‘Hi! [you are]!’
78
Imperative and exclamatory clauses tend to have more intensity and higher pitch on
the exclamatory clausal clitic =ve, pronounced with the lengthened vowel, as shown in
(2.48-2.49).
2.48
Tsameravee!
tsame=ra=ve
come.on=ADV=EXCL
‘Come on!’
2.49
Ivyaamini antaroveee!
i=pyaameni
aNtaro=ve
3m.poss=bow big=EXCL
‘Their bows were really big!’
Rising intonation is characteristic of listing, when noun phrases, predicates or clauses
are coordinated. The intonational melodies of enumerated noun phrases in (2.50) are
associated with a rise in pitch which starts around the stressed syllable and terminates at
the right edge of the noun phrase, except for the final NP which has a falling intonation.
The stressed syllables are given in bold.
2.50
Ipankitakakiri ishinkini, imachakite, iyaniri, ivaryantsitepaye, iyokani.
i=paNki-t-ak-ak-i=ri
i=shiNki-ni
i=machaki-te
3m.A=plant-EP-CAUS-PRF-REAL=3m.O 3m.poss=corn-poss 3m.poss=beans-poss
i=kaniri
i=paryantsi-te-paye
i=koka-ni
3m.poss=manioc 3m.poss=plantain-poss-PL 3m.poss=coca-poss
‘He had them plant his corn, beans, manioc, plantains, and coca.’
Content and polar questions usually have a slightly rising intonation as well, marked
with a higher pitch at the end of the final intonation unit. Examples are provided in (2.512.54).
79
2.51
Paita yaminiri soraro notsarateki?
pai-t-a
y=amin-i=ri
soraro no=tsarate-ki?
be.called-EP-REAL 3m.A=look.for-REAL=REL police 1SG.poss=bag-LOC
‘What did the police look for in my bag?’
2.52
Ninka pokatsirika?
niNka pok-atsi=ri=ka
who come-STAT=REL=Q
‘Who is coming?’
2.53
Ihataki Vela?
i=ha-t-ak-i
3m.S=go-EP-PRF-REAL
‘Did Vela go?’
2.54
Vela
person’s.name
Pinintakotana noyaani?
pi=niNt-ako-t-a=na
no=nyaa-ni
2A=want-APPL-EP-REAL=1SG.O 1SG.poss=word-poss
‘Do you want from me my words [my language]?’
At the end of songs, a final vowel is lengthened and there is a considerable rise in
pitch.
2.6 Phonological differences among Ashéninka Perené speech communities
Phonological variation is found in two geographically distinct areas, roughly comprised
of the westerly (represented by Pampa Michi) and easterly speech communities
(represented by Bajo Marankiari). The changes in the easterly speech are more advanced
than in the speech of the more conservative westerly community. Particularly, the
difference in the surface realization of the high front vowels /i/ and /e/ is more
pronounced, with /e/ articulated more open, closer to [ε] in the easterly speech
communities. For example, ‘day’ is pronounced [ˈʃitɪne] by speakers in Pampa Michi
while speakers from Bajo Marankiari say [ˈʃεtεne]; ‘little’ is [kap̥ɪ̥ ˈʧɪɪne] in the westerly
80
speech and [kapɪ̥ ˈʧeene] in the easterly speech; ‘his magic plant’ is [ɪˈßɪŋke] and
[iˈßεŋke] respectively.
In addition, the diphthong /ai/ can be monophthongized into /e/ in easterly speech.
For example, airo ‘negative irrealis’ is consistently pronounced [eɾo] by some speakers
of Bajo Marankiari. Another example of changes in the easterly speech is the lowering of
vowel height from /o/ to /a/ in vowel-initial words; for example, ‘to grab’, ‘to tie’ is
[aiɾɪk] in Bajo Marankiari and [oiɾɪk] in Pampa Michi. The vowel shift in the easterly
language communities may be possibly motivated by their regular contact with the
speakers of adjacent Kampan languages such as Ashéninka Pajonal, Ashéninka Pichis,
and Asháninka Tambo-Ene whose vowel systems include more open /e/.
Other geographically-localized differences include stopping of the aspirated alveolar
affricate /tsʰ/, alternatively pronounced as [tʰ] by some speakers of Bajo Marankiari; for
example, tsame [tsʰame] ‘come on’ is realized as [tʰame], notsoNkahero [notsʰoŋkaheɾo]
‘I’ll finish it’ as [notʰoŋkaheɾo], notsanoiriki ‘my adornments’ as [notʰanoɪɾikɪ].
Another distinction in the easterly speech concerns the emergent loss of a two-way
phonemic contrast between the aspirated /tsʰ/ and unaspirated /ts/. The westerly speakers
distinguish between the alveolar affricates /tsʰ/ and /ts/ while many easterly speakers do
not differentiate between the two. The alveolar affricate is commonly realized as
aspirated, e.g. potsotzi ‘annato’ (tree sp.) [potsʰotsʰe̥].
Labialization of the velar stop when followed by /a/ is somewhat characteristic of
westerly speech. Examples include okaNta [okʷanta] ‘then’, Katsiriki [kʷatsʰɨɾɪkɪ] ‘a
river’s name’, oNkaNtya [oŋkʷantʲa] ‘so that’.
81
2.7 Orthography11
Ashéninka Perené has never been a subject of a dedicated investigation by linguists;
hence the lack of the writing system representing the language’s phonemic inventory.
Nevertheless, two orthographies, a linguistic and a practical one, have been in circulation
in the community. The first linguistic orthography, recommended by SIL linguists in
1987 to Ashéninka bilingual teachers, was based on Ashéninka Pichis, an adjacent
language variety (Payne 1987; Anderson 1987). The pedagogical grammar of Ashéninka
Pichis and Ashéninka Perené (Payne 1989) reproduces the earlier 29-grapheme writing
system (see Table 5). The proposed system is largely based on Spanish orthography; for
example, the velar stop /k/ is written as <c> before non-front vowels a and o, and as
<qu> before front vowels i and e; the glottal fricative /h/ is written as <j>; the alveopalatal /ɲ/ is <ñ>; the bilabial approximant is written as <v>. Other Spanish graphemes,
familiar to Ashéninka Perené speakers, are <ch> /ʧ/ and <y> /j/.
The 1987 SIL set of graphemes corresponds directly to the phonemes of Ashéninka
Pichis. This means that the SIL writing system includes letters for the phonemes that do
not exist in Ashéninka Perené such as g, my, th, vy, although they are attested in its
phonemic inventory as allophones. However, the SIL writing system considers two main
criteria: orthographic depth and consistency, and the similarity of the orthography to the
orthography of the dominant language. In particular, a distinction is made between the
aspirated and unaspirated alveolar consonants and affricates by assigning digraphs <th>
to the aspirated stop /tʰ/, <ts> to the aspirated affricate /tsʰ/ and <tz> to its unaspirated
counterpart /ts/. In addition, the palatalized counterparts of stops p ,t, k, fricatives s, h, v,
11
A more detailed discussion of the development of Ashéninka Perené orthography is in Mihas 2010a.
82
affricates ts, tz, and sonorants n, m, r are also included, which are either encoded by the
sequence of Cy such as <py>,< ty>,< qy>,< jy>,< my>, <ry>,< vy>, or by digraph
<sh>, or by symbols, already familiar to Ashéninkas such as <ch> and <ñ>. The
alphabet does not differentiate between long and short vowels, listing graphemes only for
the short ones a, e, i, o.
A practical orthography has been in the process of being developed by native
speakers in Bajo Marankiari. It was designed by Edwin Jacinto Santos, a linguist with a
Master’s degree from San Marcos University in Lima, and Daniel Bernales Quillatupa, a
bilingual teacher. Their simplified alphabet, developed for the Colegio Bilingüe EbenEzer, a private Adventist high school in Bajo Marankiari (Jacinto Santos 2004), was
supported by the tribal governance unit of Area de Educación, Comunicación y Turismo
de Comunidad Indígena Asháninka Marankiari Bajo (CIAMB) (Jacinto Santos, Bernales
Quillatupa 2007). It includes 20 graphemes (see Table). This more streamlined system
disposes of <tz> and <th> and also does away with a number of other digraphs which
encode palatalized consonants such as <py>,< qy>, <my>,< ry>,< vy>. The palatalized
phonemes are now represented by Ci sequences e.g. pi, mi, ri, bi. The Spanish graphemes
<c> and <qu> are replaced with <k>. This writing system encodes /w/ as <b> and
retains grapheme <g>. In a recent attempt to support indigenous literacy, a slightly
revised alphabet (see Table 5) has been tested, which gets rid of <g> while keeping the
rest of the graphemes intact (Cushimariano Romano & Richer Sebastián Q. 2008). The
proposed writing conventions are used in the Asháninka dictionary, available online at
http://www.lengamer.org.
83
The orthographic conventions used in this grammar make a compromise between the
linguistic orthography proposed by SIL linguists and the practical orthography developed
by language speakers. With the principles of consistency in the representation of
Ashéninka Perené phonology and of user-friendliness in mind, the following solutions are
proposed. Table 5 shows the four writing systems, including the one used in this
grammar, which have been developed for Ashéninka since 1980s.
TABLE 5. Ashéninka writing systems developed from the 1980s to date
Phoneme
i
e
a
o
p
pʲ
t
tʰ
tʲ
k
kʲ
g
s
ʃ
ʧ
tsʰ
ts
h
m
n
ɲ
ɾ
ɾʲ
w
j
SIL
(Payne 1987,
1989)
i
a
o
p
py
t
th
ty
c, qu
qy
g
s
sh
ch
ts
tz
j
m
n
ñ
r
ry
v
y
CIAMB
(Jacinto,
Bernales 2007)
i
e
a
o
p
pi
t
Online version
(Romano,
Sebastián 2008)
i
e
a
o
p
pi
t
This grammar
i
e
a
o
p
py
t
ty
k
ki
g
s
sh
ch
ts
ts, t
j
m
n
ñ
r
ri
b
y
ty
k
ki
ty
k
ky
s
sh
ch
ts
ts, t
j
m
n
ñ
r
ri
b
y
s
sh
ch
ts
tz
h
m
n
ny
r
ry
v (or w)
y
N
84
As Table 5 shows, the problematic issues concern representation of palatalization and
aspiration, of the bilabial approximant /w/, and of the archiphoneme N. As far as the
grapheme N is concerned, it is used to represent morphemic structure in cases of
morphophonological changes. For example, the irrealis prefix when it precedes the
plosive-initial verb root, can take the form of [n], [m], [ŋ] depending on the adjacent
plosive. When the graphemes m and n are used to represent the above allophones, one of
the allophones [ŋ] is ignored in this shallow orthography.
To indicate orthographically palatalization, realized on stops, fricatives, a nasal, and a
liquid, two choices are available. The first option is to use a sequence <Cy>, e.g.
pyarentsi ‘manioc beer’, tyaapa ‘chicken’, avavakya ‘we’ll [let’s] eat together’, nyaantsi
‘word’, nonaryaka ‘I lay down’. The second option is to use the sequence <Cy> with the
alveolar stop t, the sequence <Ci> with stops p, k and liquid ɾ, and utilize the Spanish
grapheme <ñ> to mark the palatal nasal ɲ, e.g. piarentsi ‘manioc beer’, avavakia ‘we’ll
eat together’, ñaantsi ‘word’, nonariaka ‘I lay down’. The first approach enables
maintenance of consistency in the representation of palatalization and an adequate
expression of words whereas the second approach is inconsistent (we apply <Cy> with
the alveolar stop but with bilabial and velar stops use <Ci>; to mark palatalization of the
nasal stop, a separate symbol is used) and does not represent the words effectively.
Aspiration could be marked on the alveolar affricate by using different digraphs, <ts>
for aspirated affricate and <tz> for unaspirated one, e.g. otsitzi ‘dog’, tzimatsi ‘there is’.
Alternatively, the same digraph <ts> can be used for both, e.g. otsitsi, tsimatsi. The third
option is to use the grapheme <t> for the unaspirated affricate, e.g otsiti, timatsi. When
asked to write down recorded narratives and conversations of fellow speakers, Ashéninka
85
Perené language consultants used these techniques intermittently. Considering the
available options of marking aspiration, the first solution seems more attractive since it
avoids ambiguity, which will inevitably present a challenge to the readers who are not
proficient in the language, by assigning different digraphs to different phonemes. The
second and third solutions, although simpler, are likely to confuse a non-fluent reader and
writer, and do not adequately represent spoken words.
Finally, this grammar uses <v> to write [w] and [ß̞]. Unlike other proposals that use
<b> to represent the bilabial approximant, this study’s recommended <v> is closer to the
representation of the sound produced by native speakers. Ultimately, these problems have
to be worked out with the community and with the community’s needs in mind for the
practical orthography to function effectively. Perhaps as more Ashéninka Perené speakers
become familiar with the proposed writing conventions, these issues will be resolved
through writing workshops and informal discussions within the community.
2.8 Morphophonemics
The phonological processes discussed in this subsection deal with lenition (§2.8.1),
epenthesis (§2.8.2), deletion (§2.8.3), and sibilant contrast (§2.8.4), which occur at the
boundaries between a prefix and a root, a root and a suffix, a root and a root, and a suffix
and a suffix. Some of these phonological processes are found in other Kampan varieties
such as Ashéninka Apurucayali (Payne 1981:107-152), Ashéninka Pichis (Payne 1983),
Nanti (Michael 2008:227-8, 230-1), Caquinte (Swift 1988:107-23).
86
2.9.1 Lenition
The initial plosive element of the bilabial /p/ and velar /k/ in nouns is lost before vowels,
resulting in the plosives becoming semivowels /w/ and /j/ respectively. It occurs when
the possessive proclitics 1SG no=, 2SG pi=, 3SG masculine i=, 3SG non-masculine o=,
1PL a= adjoin the plosive-initial noun root. Examples are provided in (2.55).
2.55
LENITION OF P→W, K→J
pitotsi ‘canoe’
pava ‘god’
koka ‘coca’
kaniri ‘manioc’
kami ‘rotten leaf’
‘my_’
novito
novavani
noyokani
noyaniri
noyamini
‘your _’
pivito
pivavani
piyokani
piyaniri
piyamini
‘his_’
ivito
ivavani
iyokani
iyaniri
iyamini
‘her_’
ovito
ovavani
yokani
yaniri
yamini
‘our_’
avito
avavani
yokani
yaniri
yamini
When the velar stop is followed by the high front vowel, the lenition process results in
the loss of the velar plosive, as shown in (2.56).
2.56
LENITION OF K→Ø
‘my_’
kitsarentsi ‘clothes’ noitsari
kitzitsi ‘leg’
noitzi
kishitsi ‘hair’
noishi
kipatsi ‘land’
noipatsite
kitsoki ‘egg’
noitsoki
‘your_’
piitsari
piitzi
piishi
piipatsite
piitsoki
‘his_’
iitsari
iitzi
iishi
iipatsite
iitsoki
‘her_’
oitsari
oitzi
oishi
oipatsite
oitsoki
‘our_’
aitsari
aitzi
aishi
aipatsite
aitsoki
The lenition process is also found at the root-root boundary in noun compounds. An
example is provided in (2.57). In (2.57), the nominal stems -vatsa ‘meat’ and -pori- ‘foot,
muscle’ are combined into the noun compound -vatsa-vori- ‘meat-muscle’, which
follows the verbal stem shiriNka ‘grow’. The plosive /p/ in pori ‘muscle’ has undergone
87
the process of lenition, becoming semivowel /w/ after the word-final vowel /a/ in vatsa
‘flesh, meat’.
2.57
Ishirinkavatsavori.
i=shiriNka-vatsa-pori-t-an-ah-i
3m.S=grow-flesh-muscle-EP-DIR-REGR-REAL
‘He grew back muscle in his flesh.’
2.9.2 Epenthesis
Vowel addition.The language phonotactics does not allow consonant clusters syllableinternally and across syllable boundaries. To conform to this constraint, the epenthetic
segment -a- is used on the syllable boundary between two consonants, as shown in (2.582.59).
2.58
Okantavityata.
o=kaNt-a-vi-t-ia=ta
3n.m.S=happen-EP-FRUS-EP-IRR=OPT
‘It won’t matter.’
2.59
Nirayetzi.
n=ir-a-ye-tz-i
1SG.S=drink-EP-DIST-EP-REAL
‘I drink each of them.’
Consonant addition. The language phonotactics does not allow onsetless syllables wordinternally. To comply with this constraint, the epenthetic element t is used
intervocalically, as shown in (2.60).
2.60
Kameetsatakerika.
kameetsa-t-ak-e=rika
good-EP-PRF-IRR=COND
‘Whether it is good [or not].’
88
The epenthetic segment /t/ has two allophones, t [t] and tz [ts], the latter found
before the high front vowel [i] in realis status suffix –i. Examples are given in (2.612.62).
2.61
Ihatzi pareniki.
i=ha-tz-i
pareni-ki
3m.S=go-EP-REAL river-LOC
‘He went to the river.’
2.62
Inevetaro ihate pareniki.
i=neve-t-a=ro
i=ha-t-e
3m.A=like-EP-REAL=3n.m.O 3n.m.S=IRR-go-IRR
‘He liked to go to the river.’
pareni-ki
river-LOC
2.9.3. Vowel deletion
The vowels of the possessive proclitics 1SG no=, 2SG pi=, 3SG masculine i=12, 3SG
non-masculine o=, 1PL a= are deleted when they are adjoined to vowel-initial noun
roots, as seen in (2.63).
2.63
VOWEL LOSS I, O, A→Ø AT THE PROCLITIC-ROOT BOUNDARY
irantsi ‘blood’
oketsi ‘eye’
aikintsi ‘tooth’
‘my_’
nirani
noki
naiki
‘your_’
pirani
poki
paiki
‘his_’
irani
[iroki]
[iraiki]
‘her_’
irani
oki
aiki
‘our_’
irani
oki
aiki
Similarly, vowels of personal proclitics, which are identical with the possessive
proclitics, are elided when the vowel-initial verb root follows them. Examples with verb
roots a ‘take’, ooya ‘wait’ are provided in (2.64-2.65).
12
The 3SG.masculine possessor proclitic exhibits allomorphy i=~ir=Ø. The distribution of the allomorphs
is as follows: i= occurs with C-initial roots whereas ir= appears before a/o-initial roots. When the root is iinitial, zero marking occurs.
89
2.64
VOWEL LOSS NO →N AT THE PROCLITIC-ROOT BOUNDARY
Naakiro.
no=a-ak-i=ro
1SG.A=take-PRF-REAL=3n.m.O
‘I took it.’
2.65
VOWEL LOSS PI →P AT THE PROCLITIC-ROOT BOUNDARY
Pooyahena.
pi=ooya-ah-e=na
2A=wait-REGR-IRR=1SG.O
‘Wait for me!’
Conversely, the vowel-initial noun and verb roots may be reduced when they are
adjoined to the person or possessor markers. As seen in (2.66), the vowels which occur at
the beginning of the noun root are merged with the vowels of possessive proclitics. As a
result of this fusion, the long vowel /i:/ of the nominal root becomes short as in (2.66)
while short vowels /a/ and /i/ are elided in (2.67).
2.66
VOWEL SHORTENING I:→I AT THE RPOCLITIC-ROOT BOUNDARY
iitontsi ‘head’
iishintsi ‘hair’
2.67
‘my_’
‘your_’
‘his_’
‘her_’
‘our_’
noito
noishi
piito
piishi
iito
iishi
oito
oishi
aito
aishi
VOWEL LOSS A→Ø, I→Ø AT THE RPOCLITIC-ROOT BOUNDARY
akiri ‘nose’
ikentsi ‘neck’
nokiri
nokentsi
pikiri
pikentsi
ikiri
ikentsi
okiri
ikentsi
akiri
akentsi
2.9.4 Sibilant neutralization
The sibilants alveolar /s/ and palatal /ʃ/ are in complementary distribution when they
interact with the vowel-initial units on the syllable boundary. When the /s/-final root is
followed by high front vowels /i/ or /e/, the alveolar sibilant becomes a palatal sibilant /ʃ/.
This context-dependent variation involves neutralization with respect to place of
90
articulation, interpreted here as a process of palatalization of an underlying /s/.The
sibilant-final verb roots avis~avish ‘pass’ and pos~posh ‘hit’, exemplified in (2.68-2.71),
have only one representation in citation speech, namely avis ‘pass’ and pos ‘hit’.
2.68
Avisaki osheki osarentsi.
avis-ak-i
osheki osarentsi
pass-PRF-REAL many year
‘Many years passed.’
2.69
Pavishi.
p=avish-e
2S=pass-IRR
‘Enter!’
2.70
Oposakina tampyaa.
o=pos-ak-i=na
taNpyaa
3n.m.A=hit-PRF-REAL=1SG.O wind
‘I was hit by the wind.’
2.71
Noposhiro mapi.
no=posh-i=ro
mapi
1SG.A=hit-REAL=3n.m.O stone
‘I struck the stone.’
91
Chapter 3
Morphology
This chapter focuses on the morphology of Ashéninka Perené. The distinction between
phonological and grammatical word in Ashéninka is discussed in §3.1. A brief sketch of
morphological typology is provided in §3.2, followed by an account of free and bound
morphemes, including roots, affixes, clitics, and particles in §3.3. A detailed outline of
word classes is provided in §3.4.
3.1 Grammatical word vs. phonological word in Ashéninka Perené
Morphology is the study of the internal structure of words (Haspelmath 2002:1). There is
a need to define the notion of word first before I discuss Ashéninka morphological
structures. Two types of words will be considered, grammatical and phonological ones. It
is important for grammar-writing and lexicographic purposes to distinguish between the
two types of word.
The phonological word is established on phonological criteria. A phonological word
must have “at least one property... from the following areas: segmental features,
including word boundary phenomena; prosodic features such as stress and/or tone
assignment and vowel harmony; the domain of application of phonological rules” (Dixon
2010a:32). The criteria for a phonological word in Ashéninka are two-fold: it must have
at least two syllables and be subject to other phonotactic constraints (see the minimal
word requirement in §2.6 and phonotactic constraints in §2.7) and phonological rules;
and it has (largely predictable) stress (see §2.5 on stress placement). For example,
no.´vo.ri.ki no=pori-ki [1SG.poss=thigh-LOC] ‘from my thigh’ has second-syllable
92
stress, satisfies the minimal word requirement (it has 4 syllables), is subject to the
morphophonemic weakening rule which states that the initial plosive in nouns becomes
semivowel /w/ when possessive proclitics adjoin the noun root. Additionally, we can
recognize novoriki as a phonological word on the basis of small pause occurring at its
right boundary when it is found in a clause (the boundary is indicated by //), as seen in
(3.1).
3.1
Novoriki // yaaitakina haka.
no=pori-ki
y=a-ai-t-ak-i=na
haka
1SG.poss=thigh-LOC 3m.A=take-IMP.P-EP-PRF-REAL=1SG.O here
‘From my thigh, they took from here.’
Note that literate consultants tend to insert a space between grammatical words. Curiosly,
they write down reduplicated verb roots as separate phonological words rather than string
them together as parts of one grammatical word, as shown in (3.2).
3.2
*Okiva kivatzi.
o=kivakiva~tz-i
3n.m.S=wash.a.lot~EP-REAL
‘She was washing a lot of clothes.’
The second type of word, a grammatical one, is the basic unit for grammatical
operations. Grammatical words in Ashéninka Perené are generally composed of strings of
concatenated morphemes, which include roots and layers of formatives (affixes, clitics,
particles). For the purpose of the grammatical description of the language, the
grammatical word is the most relevant since it serves as a basic unit of morphological
description. By definition, a grammatical word is established on grammatical criteria and
is the product of morphological process (e.g. compounding, reduplication, shift of stress,
93
affixation, etc.), applied to its lexical root(s) to express a conventionalized meaning
(Dixon 2010a:13). In Ashéninka Perené, grammatical words are defined on
morphological and syntactic criteria. A morphological property restricted to a certain
word class found in the language serves as the morphological criterion for a grammatical
word. This criterion may apply with respect to inflections on verbs and nouns. For
example, aspectual distinctions such as durative –vai, progressive –aty~-atz, perfective –
ak, etc. are restricted to verbs only exemplified by (3.3).
3.3
Pinyaavaite.
pi=nyaa-vai-t-e
2S=speak-DUR-EP-IRR
‘Speak!’
Case is an example of morphological category, largely restricted to the noun class.
The polyfunctional locative case marker –ki is found with nouns (or associated with the
noun elements), as given in (3.4).
3.4
Nakoki karakina.
n=ako-ki
kar-ak-i=na
1SG.poss=arm-LOC break-PRF-REAL=1S.O
‘I broke my arm.’
By making reference to the word-class-specific morphological category, these
morphological forms serve as evidence for the notion of grammatical word. Some
morphological processes can also be word-class-specific, e.g. reduplication is limited to
the verbs and onomatopoeic forms. The syntactic criterion for independent grammatical
words is based on distribution of grammatical words. In (3.5), an adverb is discontinuous
with the verb.
94
3.5
Pairani nocharinite ipaitziro nihaa Meantarini.
pairani no=charini-te
i=pai-tz-i=ro
long.ago 1SGposs= grandpa-poss 3m.A=call-EP-REAL=3n.m.O
nihaa Meantarini
river name
‘Long ago our ancestors called the river Meantarini.’
Additionally, the writing habits of literate language consultants generally indicate
what type of word we are dealing with; they tend to insert spaces around a grammatical
word, although in some cases they place word boundaries around a larger phonological
unit. For example, the phonological word poo (which violates phonotactic constraints on
a phonological word), an ideophone imitating a hitting sound is often written attached to
the noun whose referent indicates the area of penetration as seen in (3.6).
3.6
*Noitokipoo!
no=ito-ki
poo
1SG.poss=head-LOC IDEO
‘(It hit) my head, bam!’
In general, speakers tend to make small pauses between grammatical words (which
often coincide with phonological words) but not within a word.
Examples of grammatical words in Ashéninka Perené will be inflected verbs and
nouns as well as unbound nominal roots which constitute the core of the grammatical
word, with formatives attached to the root serving as parts of the grammatical word. For
example, a free unpossessed noun root pava ‘god’ is a grammatical word which derives a
possessed noun ivavani i=pava-ni [3m.poss=deity-poss] ‘their god’. A repeated
onomatopoeic word form tsikiri-tsikiri ‘sound of liquid gushing out in a stream’ is also a
grammatical word and so is the noun compound incha-tonki [plant-long.segment] ‘wood
95
plank’ which consists of two noun roots13. The negative particle te and the interjection
akayave, used to express pain, are examples of independent grammatical words; others
are free pronouns, demonstratives, locative adverbs, interrogatives, adverbs.
In the majority of cases, phonological and grammatical words coincide in Ashéninka
Perené. However, there may be a few mismatches. For example, the compound noun is a
grammatical word which includes two phonological words as in the class term 'in.cha'ko.ta [plant-piece] ‘split wood’ where stress applies to each part of the nominal
compound. Conversely, a single phonological word can consist of two grammatical
words. For example, te oniNte te'niNt-e [NEG.REAL.want-IRR] ‘she didn’t want’ is
realized in fast speech as one phonological word, with the negative particle te cliticizing
to the verb. As a result, two grammatical words fuse into a single phonological word. The
stress serves as the litmus test for the identification of a phonological word. Note that the
negative particle te routinely forms one phonological word with vowel-initial predicates.
3.2 Morphological typology
Morphological typology is based on the extent to which words of a given language are
divisible into “clearly individuated morphemes” (Payne 1997:27). Ashéninka Perené is a
highly polysynthetic language whose words tend to consist of multiple morphemes.
Polysynthesis involves internal morphological complexity when large numbers of lexical
and grammatical morphemes are combined within one word (Aikhenvald 2007:5). On the
13
The N+N compounds (e.g. incha-tonki ‘wood plank’) and N+CL (e.g. kanai-ki [Long John tree-CL:
social.insect] ‘insect living in the Long John tree’ combinations are distinguished on the basis of
distributional and structural properties, and semantics of their parts. The compound parts can occur
independently in discourse; each part is semantically transparent and does not undergo loss of phonological
substance. The second elements of the N+CL combinations need a host to be integrated into discourse; they
typically exhibit bleached semantics and phonological reduction.
96
basis of Aikhenvald’s (2007) definitional criteria of polysynthesis, the following traits
characterize polysynthesis in Ashéninka Perené.
(i) Noun incorporation, as illustrated by (3.7).
3.7
Amamenkotziyetaki.
ama-menkotzi-ye-t-ak-i
3n.m.S.swim-disheveled.head-DIST-EP-PRF-REAL
‘Many heads floated [in the river].’
(ii) A large inventory of bound morphemes.
(iii) Extremely productive derivational processes.
(iv) Polypersonalism, that is cross-referencing of A, S, O and sometimes other
arguments on the verb and possessors on the nominal form.
(v) Integration of adverbial (manner, time, direction, degree) elements into the verb
complex, as shown in (3.8-3.9).
3.8
Haitetaki.
ha-ite-t-ak-i=Ø
go-quickly-EP-PRF-REAL=3n.m.O
‘She went quickly.’
3.9
Ikemakanintari.
i=kem-aka-niNt-a=ri
3m.A=listen-CAUS-DIM-REAL=3m.O
‘He made him listen a little bit.’
(vi) Many possible affixal slots, with a few of them obligatory.
Polysynthetic languages are often agglutinating, so is Ashéninka Perené. Its
morpheme boundaries are clearly identifiable, with little allomorphic variation. Each
morpheme typically has one meaning. However, there are a few exceptions exemplified
by fused portmanteau inflections such as first and third person clitics no= 1SG, a=1PL,
97
=na 1SG, =ai 1PL, i= 3m, o= 3n.m, =ri 3m and =ro 3n.m. which refer to more than one
grammatical category (person, number, or gender).
A typical property of polysynthetic languages is head-marking: syntactic relations are
marked on the head rather than on the dependent. The head is the element that determines
the syntactic function of the whole phrase (Payne 1997:31). The marking of the possessor
on the noun is an example of head-marking as in novaNko no=paNko [1SG.poss=house]
‘my house’. Another instance of head marking such as cross-referencing of the core verb
arguments on the verb and manipulation of other arguments through applicatives and
classifiers, also marked on the verb, is illustrated by (3.10-3.11) respectively.
3.10
Ikantziro.
i=kaNt-tz-i=ro
3m.A=say-EP-REAL=3n.m.O
‘He said to her.’
3.11
Okoyekitaki.
o=koye-ki-t-ak-i
3n.m.S=gather-CL:small.round-EP-PRF-REAL
‘She has harvested coffee beans.’
.
Morphological processes in Ashéninka Perené include cliticization, prefixation,
suffixation, infixation, reduplication, compounding, and noun incorporation which will
be addressed in detail in §3.4.
3.3 Formative classes
3.3.1 Definitions of free and bound morphemes, roots, formatives
In this section, I will briefly discuss some terms used in this account of Ashéninka
morphology. Morphemes are defined as “the smallest meaningful constituents of
98
linguistic expression” (Haspelmath 2002:16). Words are multimorphemic in Ashéninka.
For example, the word nonoshikakemirika ‘if I pull you’ consists of six meaningful units
or morphemes: the verb root noshik ‘pull’, two phrasal personal clitics no= ‘1SG.S’ and
=mi ‘2O’, clausal conditional enclitic =rika, and two affixes –ak ‘perfective aspect’ and
–e ‘irrealis status’. Free morphemes are realized as phonologically and syntactically
independent units. Free morphemes are unpossessed nouns, adjectives, adverbs, clause
connectors, demonstratives, pronouns, interjections, ideophones, and particles. Bound
morphemes are always morphological entities that must be attached to some other
morpheme in order to be integrated into discourse (Payne 1997:21). Roots, affixes, and
clitics are examples of bound morphemes. In polysynthetic languages like Ashéninka
Perené, there are more bound than free morphemes. Bound morphemes are all verbs,
obligatorily possessed nouns, classifiers, affixes, and clitics. Clitics are represented by the
equation sign = while other bound elements are indicated by hyphens, e.g. classifier –ki
‘small, round’ as in tsoiro-ki ‘snail species’; adjectivizer -ri as in santsa-ri [be.thin-ADJ]
‘thin’; causative prefix v-, directional, perfective, and realis suffixes –ap, -ak, -i
respectively, and personal clitics i= ‘3m.S’and =na ‘1SG.O’ as in Ivaniitapakina
i=v-anii-t-ap-ak-i=na [3m.A=CAUS-go.down-EP-DIR-PRF-REAL=1SG.O] ‘He took me
down’. Affixes are positioned closer to the root while clitics appear on the outside
boundaries of the word.
Root is defined as “an unanalyzable form that expresses the basic lexical content of
the word” (Payne 1997:24). There are unbounded and bounded roots in Ashéninka.
Unbounded roots are at least disyllabic and form independent grammatical and
phonological words as in ari ‘positive polarity verb’, kametsa ‘be good’, iri ‘he’. Verbal
99
roots are always bound. Examples of monosyllabic verb roots are na ‘to be’, ny ‘to see’,
nya ‘to speak’, (o)v ‘to eat’, ‘to kill’, p ‘to give’, a ‘to take’, ir ‘to drink’. Examples of
disyllabic roots are tsarov ‘to fear’, shitov ‘to run/escape’, nimo ‘to like’. Some
polysyllabic roots appear to be derived from other forms as in kiNkitsa ‘to tell’, kamatsa
‘to crawl (on the floor)’, kimisaNt ‘to pay attention’, apishoiNk ‘to throw’. Nominal roots
that are obligatorily possessed are always bound, e.g. oito ‘head, yeNpita ‘ear’, paNte
‘mouth’, kishi ‘hair’, tzinko ‘tail’, yeNka ‘fat’. Examples of unbound nominal roots are
mapi ‘stone’, ani ‘brother-in-law’, maini ‘bear’, shiro ‘dove’, nihaa ‘water’, tzivana
‘pineapple’, chamaNto ‘woodpecker’. Examples of independent adjectival roots,
adverbial roots, and function words are kiNkivaro ‘old’, iyaani ‘small’, kapeechi ‘little’,
osheki ‘many’, hiri ‘here’, pairani ‘long ago’, ironyaaka ‘now’.
Many roots are polyfunctional, i.e. one form can belong to different lexical
categories. For example, verbal root shintsi ‘be strong’ is an adverb in (3.12), verb in
(3.13), noun in (3.14), and adjective in (3.15).
3.12
Shintsini tekirata ontsiteniteta.
shintsi-ni
tekira=ta
be.strong-ADV not.yet=OPT
‘Hurry up before it gets dark!’
o=N-tsiteni-t-e=ta
3n.m.S=IRR-dark-EP-IRR=OPT
3.13
Pishintsitahatyera.
pi=shintsi-t-ah-aty-e=ra
2S=be.strong-EP-REGR-PROG-IRR=ADV
‘Hurry up, will you?’
3.14
Evankari tzimatsi ishintsinka.
evaNkari
tzimatsi i=shintsi-Nka
young.man EXIST 3m.poss=be.strong-NMZ
‘The young man has strength.’
100
3.15
Shintsiri okanta shivitsa.
shintsi-ri
o=kaNt-a
shivitsa
be.strong-ADJ 3n.m.S=be-REAL rope
‘The rope is strong.’
Formatives are morphemes which do not head phrases, govern or are governed, or
trigger/undergo agreement (Bickel &Nichols 2007:172-3). In Ashéninka Perené,
formatives include affixes, clitics and particles. A distinction can be made between core
and peripheral formatives, following the definitional criteria proposed by Epps
(2006:166-7). The two types of formatives are identified on the basis of their distance
from the root and the degree of their phonological integration with it. The core formatives
in Ashéninka include affixes while peripheral formatives are made of clitics and particles.
As far as their distance to the root is concerned, core formatives are positioned closer to
the root while peripheral forms appear on the outside boundaries of the word; peripheral
forms (clitics and particles) are always external to the core formatives (affixes). The
verbal template is summarized below.
Particle
Proclitic=
Prefix-
ROOT
-Suffix
=Enclitic
The nominal template typically corresponds to the following.
Proclitic=
ROOT
-Suffix
=Enclitic
When it comes to the degree of their phonological integration with the root, the core
formatives tend to bear stress and participate in morphophonemic processes. In contrast,
peripheral formatives are generally not stressed. However, because these two groups of
formatives are themselves internally heterogeneous, the distinction between them is
101
gradient, which makes challenging, in particular, the classification of person and
possessor morphemes.
This analysis does not include the derivational or inflectional functions of Ashéninka
formatives as a definitional criterion. Arguably, derivational morphology, though nonobligatory, is more relevant to the concept expressed by the root, which motivates its
close proximity to it. In contrast, inflectional operations are often grammatically required
in order to create fully-formed words but are less relevant to the root, hence their
peripheral placement on the word boundary (Payne 1997:25-6). Ashéninka does not
appear to abide by this cross-linguistically common ordering principle. In Ashéninka,
some inflectional formatives can be found both next to the root, e.g. plural number
marker –het, and on the periphery of the verbal word, like aspect and realis status
markers -ak ‘perfective’ and –i ‘realis status’, as shown in (3.16).
3.16
Isaikahetapaki Metaroki.
i=saik-a-he-t-ap-ak-i
Metaro-ki
3m.S=be.in-EP-PL-EP-DIR14-PRF-REAL village’s.name-LOC
‘They all settled down in Metraro.’
Associated with inflectional morphology, the criterion of obligatoriness of the
formative is not used here either, on the grounds that the obligatory forms in Ashéninka
include both core and peripheral formatives. For example, a well-formed verb has to be
minimally inflected for subject and reality status as in Piri p=ir-e [2S=drink-IRR] ‘Have
14
In this analysis, directionals –ap ‘goal’, -an ‘source’, -av ‘O-oriented goal’ are placed on the continuum
from inflection to derivation as ‘intermediate’ items since they have both derivational and inflectional
properties. On the one hand, directionals are non-obligatory; on the other hand, the directionals –ap and
–an, in addition to their basic directional senses, express aspectual meanings of perfectivity (‘do something
on arrival’) and ingressivity (‘begin doing something’) respectively; the directional –av essentially codes a
semantic patient in the proposition.
102
a drink!’ that comprises the S/A cross-referencing proclitic p= , a peripheral formative,
and the realis status suffix –i which appears on the core boundary.
The next section will describe affixes, which is a heterogeneous group of formatives,
closely associated with the root.
3.3.2 Core formatives: affixes
Affixes are defined as “short morphemes with an abstract meaning” (Haspelmath 2002:
18). Ashéninka Perené is mainly suffixing, with only a few prefixes, which is typical of
many Arawak languages (Aikhenvald 1999:80). There are only two prefixal slots that
precede the root, as opposed to a large number of formatives that follow it. Suffixes tend
to occur in a fixed order, but some may be found in a different position on the verb
template. There are about 40 verbal suffixes and about 39 nominal suffixes, and most of
them are category-specific. Affixes may be monosyllabic or have a di- or trisyllabic
phonological form. There are also a few infixes and two discontinuous affixes
(circumfixes).
Prefixes are exclusively verbal. The group of prefixes includes three members, the
irrealis status N- and the causative prefixes malefactive mi- and causative o-~oi-~ov-. The
causative prefixes are valence-adjusting and the irrealis status prefix refers to future
temporality. Speakers disprefer to use both irrealis prefix N- and causative prefixes in the
same slot and omit the prefixal irrealis status marker, indicating irrealis status by the
irrealis suffix only, as seen in (3.17-3.18).
103
3.17
Ivaniitapakina.
i=v-anii-t-ap-ak-i=na
3m.A=CAUS-go.down-EP-DIR-PRF-REAL=1SG.O
‘He took me down [when I was unconscious].’
3.18
Ari ivaniitahena.
ari i=v-anii-t-ah-e=na
PP 3m.A=CAUS-go.down-EP-REGR-IRR=1SG.O
‘He will take me down.’
Verbal suffixes in Table 6 include verbal number, valence-adjusting/preserving,
adverbial (manner, direction, time), aspectual and mood formatives. Some of them can
occur with nouns and some with other word classes. For example, –he ‘plural number’,
classifier –ki ‘small, round’, classifier -poroki ‘bunch’, group’ are found with nouns as
well. Verbs minimally require a reality status suffix when functioning as predicates.
The identification of the suffix zones in Table 6 is made on the basis of broad
generalizations relating the formative class to its semantics and function. The indicated
suffix zones include plural verb number, classifier, valence-adjusting/ preserving,
adverbial, and aspectual/status suffix subsets.
In (3.20), the string of suffixes is comprised of the causative –ak, directional –an,
perfective aspectual –ak, and realis status –i morphemes, the last two assigned to the
same boundary suffix zone.
3.20
Ikinakanakina intakironta.
i=kin-ak-an-ak-i=na
iNtakiro=Nta
3m.A=pass-CAUS.SOC-DIR-PRF-REAL=1SG.O inside=ADV
‘He walked with me inside over there.’
104
TABLE 6. Verbal suffixes
Zone
Plural number suffix
Classifier suffixes
Valence-adjusting or
valence-preserving suffixes
Adverbial suffixes
Boundary suffix zone
(aspectual, realis status
suffixes)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
Form
-he (…-ni)
-ki
-poroki
-ako
-shi
-ak
-ako
-veNt~-viNt
-imo
- pitsa
-aNt
-ashi
-av
-ry
-ai
-ye
-pero~-piro
-ve~-vi
-ap
-an
-av
-niNt
-aman
-ite
-it
-apaiNt
-apanaNt
-a
-it
-ah
-ak
-ima
-aNt
-ap(i)iNt
-vai
-aty~-atz
-atsi~-acha
-ai(n)ts
-i/-a
-e/-ia
Function with verbs
plural number
classifier ‘small, round’
classifier ‘quantity’
classifier ‘vessel’
classifier ‘small, thin, flexible’
sociative causative
generalized applicative
beneficial applicative
presential applicative
separative applicative
instrumental/reason applicative
intent applicative
reciprocal
reversative
impersonal passive
distributive
intensifier ‘indeed’ (degree)
frustrative
directional goal
directional source
directional O-oriented goal
almost, a little bit (degree)
early
quickly
temporal anteriority marker
punctual (‘once’) aspect
iterative (‘one more time’)
repetitive aspect (‘again and again’)
incompletive (circumstantial) aspect
regressive/terminative aspect
perfective aspect
inchoative aspect
customary aspect
habitual aspect
durative aspect
progressive aspect
stativity marker
temporal stativity marker
realis status
irrealis status
105
Although the rationale for assigning suffixes to a certain zone is largely determined by
their semantics and function (e.g. formatives related to the valence mechanism are
grouped together), each subset of formatives is heterogeneous, and in some cases it is
difficult to predict their function on the basis of their placement within a certain zone. For
example, the derivational suffix –ve ‘frustrative’ follows the aspectual suffix –vai
‘durative’ in (3.21).
3.21
Iposavaivitakana.
i=pos-a-vai-ve-t-ak-a=na
3m.A=hit-EP-DUR-FRUS-EP-PRF-REAL=1SG.O
‘He kept slapping me [on the face but I didn’t wake up].’
The frustrative suffix is expected to fill the position closer to the verb root but instead the
durative aspect morpheme occurs first in the string15. Note that the placement of the
formatives in the same zone does not preclude them from appearing in the same slot
together. For example, the causative and instrumental suffixes in (3.22) are from the same
valence-adjusting/preserving subset.
3.22
Pishiyakantapaintero.
pi=shiy-ak-aNt-apaiNt-e=ro
2A=be.similar-CAUS-APPL.INST-once-IRR=3n.m.O
‘Take a picture.’
Nominal suffixes comprise of classifying, plural, distributive, locative, diminutive,
possessive, and nominalizing morphemes, as seen in Tables 7-9.
15
Such morphemes, exhibiting fluidity in their positioning within the zone, are not infrequent in Ashéninka
varieties. For example, Payne notes that within the core formatives zone, in which up to five elements may
occur, “the order of these elements is quite variable. I have witnessed on numerous occasions native
speakers of Asheninka employing a verb with two or three suffixes…, switching the order of the suffixes
around on successive repetitions of the verb” (2002: 491-2).
106
TABLE 7. Nominal suffixes16
No
1
Form
-ri~ro
2
3
4
5
6
see Tables below
-ki
-paye
-te,-ni,-ri
-rintsi~-rontsi,
-(me)Nto,-Nka, -(n)tsi,-ri
-ite, -mashi, -saNte
7
8
-patsaini, -(i)ni, -aniki,
-peta
Function
gender agreement
suffix
classifying morphemes
locative case
plural number
possessive marker
nominalizer
augmentative
diminutive
Examples
kishiri ‘comb’
naNpitsi-ki ‘in the village’
kooya-paye ‘women’
imachakite ‘his beans’
saNkinarentsi ‘book’
itsaroNka-ki-ni ‘in their fear’
ivaNkoite ‘his big house’;
oparayamashi ‘big beach’
eentsi-patsaini ‘little kids’, oshiNtaniki ‘her little daughter’
TABLE 8. Class terms as derivational suffixes
No
1
Form
-to
Function
tree, trunk
Examples
inchato ‘tree’
2
3
4
5
6
7
8
9
10
11
12
-shi
-ki
-kii
-chee
-pa
-Nki
-ki
-nari
-tsori
-ni
-satzi
leaf
seed, fruit
stick
thorn, spine
plantain
snake
social insect
wild, demonic
distant kinship
deceased human referent
human provenance
inchatoshi ‘tree leaf’
kahaiki ‘coffee bean’
inchakii ‘long, thin branch’
savorochee ‘thorn of cane reed’
tyokoripa ‘white plantain’
noonki ‘boa’
imoki ‘grub’
manironari ‘mythical personage’
yayatsori ‘woman’s cousin’
chookini ‘deceased sister’
kishiisatzi ‘person from mountainous grasslands’
TABLE 9. Nominal classifiers as derivational suffixes
1
2
3
4
5
6
7
8
16
Classifier
-na
-naki
-tsa
- pi
-shi
-ki
-pa
-meNta
Semantics
long, hollow objects
round, hollow objects
long, thin, flexible objects
rigid, hollow objects
small, thin, flexible objects
small, round; hard objects
cylindrical, pod-like objects
flat, triangular objects
Examples
omorona ‘hole’
kovitsinaki ‘pot’
paamaritsa ‘power line’
chakopi ‘arrow’
novankoshi ‘my temporary shack’
potooki ‘tree sp.’
akapa ‘avocado’
inchamenta ‘hiding place in the form of a
triangle, formed by the tree roots’
Details of the Ashéninka Perené nominal classification system are provided in Mihas 2010b.
107
The relative ordering of nominal suffixes is as follows.
NOUN
ROOT
Nominalizer Classifying
morpheme
Possessive Locative
marker
marker
Diminutive/augmentative
marker
Plural marker
Examples of nominal morphology are provided in (3.23-3.27).
3.23
Tzimatsi ninchashitepatsaini.
tzimatsi n=incha-shi-te-patsaini
EXIST 1SG.poss=plant-CL:small.flexible-poss-small.part.DIM
‘I have my little herbs.’
3.24
Nontsiyakiri iitzikipatsaini.
no=N-tsiy-ak-e=ri
i=itzi-ki-patsaini
1SG.A=IRR-heat-PRF-IRR=3m.O 3m.poss=foot-LOC-small.part.DIM
‘I’ll steam him, [directing the steam] to his little foot.’
3.25
Ihatashitakiri ivankoshiki.
i=ha-t-ashi-t-ak-i=ri
3m.A=go-EP-APPL.INT-EP-PRF-REAL=3m.O
i=paNko-shi-ki
3m.poss=house-CL:small.flexible-LOC
‘He approached his shack.’
3.26
Itsaronkakini ikantahetaki.
i=tsaro-Nka-ki-ni
i=kaNt-a-he-t-ak-i
3m.poss=fear-NMZ-LOC-DIM 3m.S=be-EP-PL-EP-PRF-REAL
‘They were living with fear.’
3.27
Kitaiterikipaye.
kitaite-ri-ki-paye
be.morning-NMZ-LOC-PL
‘Over the many days [which passed].’
Adjectives and adverbs also take suffixes. The same suffix can be assigned to
multiple word classes including adjectives and adverbs. Some of the adjectival and
108
adverbial suffixes occur with nouns and/or verbs, e.g. –ni, -ki, -ri, -ro, -ra. In some cases,
their meanings overlap, in other cases they are distinct, as shown in Table 10.
Examples of adjectives and adverbs are as follows: iyaani-ni [small-DIM] ‘very
small’, osheki-ni ‘a lot’ [many-AUG] , aNtaro-ite [big-AUG]‘very big’, o=v-ak-e-ra(-ni)
[3n.m.S=place-PRF-REAL-ADV(-ADV)]‘recently’, shimaana-ki [week-LOC] ‘during
the week’, mantsiy-a-ri [be.sick-EP-ADJ] ‘sick’.
TABLE 10. Suffixes which occur with multiple word classes
Form
-ni
-ki
-ite
-ri/-ro
Function
Nouns
deceased noun
referent;
alienable
possessive
marker
locative case
augmentative
gender-sensitive
agreement
markers
-ri
-ra
medial
demonstrative
clitic
Verbs
augmentative;
plural marker
Adjectives
diminuitive
locative case
augmentative
gender-sensitive
agreement
markers;
nominalizing/relativizing adjectivizer
clitic
adverbial subordinating
clitic ‘when’; adverbial
clitic ‘where’
Adverbs
augmentative
temporal marker
quickly
person clitics
adverbializer
3.3.3 Peripheral formatives: clitics and particles
Peripheral formatives (clitics and particles) are distinguished from core formatives
(affixes) due (i) to their distribution, i.e. they are always external to core formatives, (ii)
they tend to have more mobility: tendentially, their position may not be fixed with regard
to the host or other clitics. Peripheral formatives include two particles (te ‘negative realis’
and airo ‘negative irrealis’) and clitics. There is a small class of proclitics and a larger
109
class of enclitics. Neither proclitics nor enclitics exist as independent words. By
definition, clitics are phonologically bound forms which attach to hosts of diverse
categories. Clitics are category-neutral, that is they are unrestricted as to the syntactic
category of the word they attach to (Bickel & Nichols 2007:174,176). Fixed positition
phrasal clitics (person and possessor clitics) occupy the slot adjacent to the head of the
VP or NP whereas high mobility clausal clitics attach to their hosts in different positions
in the clause. For example, demonstrative enclitics =ka, =ra, =Nta can attach to (i)
nominals, (ii) verb stems where they function as adverbials, and (iii) to function words
such as negative particle te where the demonstrative enclitic =ra has a causal sense ‘as,
since, because’. The above functions are illustrated by (3.28-3.30) respectively.
3.28
noimiporokira
no=imi-poroki=ra
1SG.poss=husband-bunch=DEM
‘that husband, bag of bones’
3.29
Osaikira...
o=saik-i=ra
3n.m.S=be.at-REAL=ADV
‘where she was’
3.30
Tera noshirityaro…
te=ra
no=shiri-t-ia=ro
NEG.REAL=ADV 1SG.A=soul-EP-IRR=3n.m.O
‘As I didn’t think about it…’
I will briefly discuss here the status of person and possessor morphemes by drawing
on the clitichood parameters proposed by Halpern (1998), Aikhenvald (2002), and
Anderson (2005). The rationale for classifying person and possessor markers as phrasal
clitics includes the following.
110
(i) The morphemes’ prosodic dependence, i.e. clitics do not bear stress and are not
counted for the purpose of stress assignment, as shown in §2.4.3.
(ii) Their external position with regard to affixes, i.e. person and possessor markers
always occur outside all suffixes (see the verbal and nominal templates in §3.3.1).
(iii) Their syntactic function, i.e. person markers function as the core verb arguments;
syntactic rules of the language are sensitive to the arguments that are coded on the verb
and apply regardless of whether the referent NPs are present or not.
(iv) As far as the ordering of clitic strings is concerned, only clausal enclitics follow
person enclitics; person proclitics are not preceded by any proclitics.
(v) Their relative selectivity (proclitics attach as subject markers to verbs, e.g. no=hate ‘I
will go’ and as possessor marker to nouns, e.g. no=shinto ‘my daughter’; they are used as
enclitics for intransitive subject marking and object marking on verbs, e.g. shitovahana
[shitov-ah-a=na] ‘I ran away’, onyapahina o=ny-ap-ah-i=na [3n.m.A=see-DIR-REGRREAL=1SG.O] ‘She saw me’).
(vi) Person and possessor markers are associated with the morphological class of personal
pronouns which tend to grammaticalize into clitics before they develop further into
affixes (see §3.4.5 for the development of pronominal bound forms in Ashéninka
Perené).
Nevertheless, proclitics show signs of affix-like behavior when they trigger
phonological adjustments in their host and undergo adjustments themselves (see §2.7 on
morphophonemic processes involving proclitics). Person enclitics in O function do not
delete when they appear in the clause along with coreferential focused pronouns (see
§3.4.5 for detailed discussion of the focused pronouns and their distribution). S/A person
111
proclitics and possessor proclitics are obligatory, thus showing more affix-like
characteristics than person enclitics, which are optional. In sum, Ashéninka Perené
person clitics are closer to the affix end of the clitic grammaticalization continuum
independent word > clitic > affix, considering that clitics are defined as procodically
deficient elements which occupy “a particular place within a multidimentional
continuum, from a fully bound to a fully independent morpheme” (Aikhenvald 2002:43).
A list of Ashéninka proclitics is provided in Table 11; enclitics are listed in Table 12.
TABLE 11. Proclitics
Host
Verbal
predicate
Form
n(o)=
p(i)=
i=,y=, ir=
o=
a=
Function
1pSG subject
2p subject
3p masculine subject
3p non-masculine
subject
1pPL subject
Other functions
1pSG possessor (nouns)
2p possessor (nouns)
3p masculine possessor (nouns)
3p non-masculine possessor (nouns)
1pPL possessor (nouns)
As Table 12 shows, enclitics perform various functions. Demonstrative enclitics
which combine with nouns are used to draw the hearer’s attention to an entity and
indicate its approximate location. Personal enclitics which attach to verbal predicates
cross-reference O arguments. Other enclitics code subordination, mood categories, or
pragmatic functions. There are four ‘floating’ pragmatic enclitics which have an emphatic
meaning when attached to any constituent in the clause. Examples of enclitics are
provided in (3.31-3.35).
112
TABLE 12. Verbal enclitics
Host
Primarily nominal
Primarily verbal
Attach
indiscriminately to
various hosts
Form
=ka
=ra
=Nta
=na
=mi
=ri
Function
proximal demonstrative
medial demonstrative
distal demonstrative
1SG.O argument
2O argument
3masculine O argument
Other functions
locative adverbial on verbs
locative adverbial on verbs
locative adverbial on verbs
=ro
nominalizer
=ni
3non-masculine O
argument
3 person recipient/
beneficiary; unspecified
for gender
=ai
=ri
1PL O argument
relativizer
=ni
=ra
relativizer
locative adverbial ‘where’
=raNki
=(tai)ma
fused time adverbial/past
dubitative
=mi
counterfactual condition
marker
=rika
=ka
possible conditional clause
linker ‘if’
interrogative
=ta
optative
=kari
undesirable possible
consequence clause linker
‘lest’
contradictory emphasis
non-contradictory
emphasis
O focus marker?
general focus marker
=ha
=tya
=kya
=ve
relativizer, adjectivizer,
nominalizer suffixes
verbal plural suffix,
augmentative suffix, nominal
possessive suffix, nominal
‘deceased referent’ marker
nominalizer suffix,
adjectivizer suffix
adverbializer
medial demonstrative on
nouns; temporal subordinating
clitic ‘when’
past tense marker on nouns
emphatic marker on nonverbal constituents?
avertive marker; marker of
frustrative modality; affective
marker on verbs
temporal clause marker ‘when’
locative adverbial on verbs;
proximal demonstrative on
nouns
emphatic marker on nonverbal constituents?
single-clause warning marker
temporal suffix ‘almost’ on
time adverbs
113
3.31
Tzimatsi ninchashipatsaini nopankitzirira.
tzimatsi n=incha-shi-patsa-ini
no=paNki-tz-i=ri=ra
EXIST 1SG.poss=plant-leaf-part-DIM 1SG.A=plant-EP-REAL=REL=ADV
‘I have herbs which I planted there.’
3.32
Novayetziri inchashi nonkaatantyariri mantsiyari.
no=v-a-ye-tz-i=ri
incha-shi
1SG.A=place-EP-DIST-EP-REAL=REL plant-leaf
no=N-kaa-t-aNt-ia=ri=ri
mantsiy-a-ri
1SG.A=bathe-EP-APPL.REAS-IRR=3m.O=REL be.sick-EP-ADJ
‘The leaves that I put here are for bathing the sick in steam.’
3.33
Tekirata yookerota yametahetari pairani.
tekira=ta
y=ook-e=ro=ta
not.yet=OPT 3m.A=throw-IRR=3n.m.O=OPT
y=amet-a-he-t-a=ri
pairani
3m.S=be.accustomed-EP-PL-EP-REAL=REL long ago
‘They haven’t yet abandoned what they were accustomed to long ago.’
3.34
Orave ikantaitzironta tzivi.
ora=ve
i=kaNt-ai-tz-i=ro=Nta
tzivi
that.n.m.=G.FOC
3m.A=say-IMP.P-EP-REAL=3n.m.O=ADV salt
‘This over there they call it salt.’
3.35
Nonkitsatavakyatatyakya.
no=N-kitsa-t-av-ak-ia=ta=tya=kya
1SG.S=IRR-dress-EP-DIR-PRF-IRR=OPT=N.C.FOC=O.FOC
‘I am going to dress [somebody].’
Ashéninka Perené enclitics often pile up. When more than one enclitic attaches to a
host, the enclitics tend to appear in a certain order. The first layer of verbal enclitics
consists of low mobility personal clitics. Their position is always fixed with respect to
their host. The second layer includes subordination and mood categories clitics, followed
by the final layer of pragmatic enclitics. The template of primarily verbal enclitics is
given below.
114
ROOT
Affixes
Personal
enclitics
Subordination enclitics
Mood categories
enclitics
Pragmatic enclitics
The ordering of nominal enclitics is as follows.
ROOT
Affixes
Demonstrative enclitics =ka, =ra,
=Nta
Past tense enclitic =raNki
Pragmatic enclitics
When attached to the nominal host, the second position demonstrative enclitics are
followed by emphatic enclitics, as seen in (3.36-3.37).
3.36
Picheckero inchapanki novanikintave.
pi=chek-e=ro
incha-paNki no=vani-ki=Nta=ve
2A=cut-IRR=3m.O plant-plank 1SG.poss=plot-LOC=ADV=EXCL
‘Please cut the wood on my land.’
3.37
Kimitaka avaniranki.
kimit-ak-a
a=vani=raNki
seem-PRF-REAL 1PL.poss=plot.of.land=ADV.P
‘It seems that this is our land [although we don’t own it now]17.’
There are two grammatical particles in Ashéninka Perené, the preverbal negative
morphemes te ‘negative realis’ and airo ‘negative irrealis’. Particles are phonologically
free units (Bickel & Nichols 2007:173). Although the negative particle te tends to
cliticize to the verb and to participate in vowel hiatus processes if the verb is vowelinitial, it may be treated as an independent phonological word since it can receive stress
when other clausal constituents can come between it and the host. When it happens,
17
The adverbial past enclitic =raNki is used on nouns to indicate that the referent of the noun no longer
exists or functions as such, or when the possessive relationship has ended.
115
pause marks the boundaries of the phonological words. Examples of the particle are given
in (3.38-3.39).
3.38
Te naaka nonpashiventya.
te
naaka no=N-pashiveNt-ia
NEG.REAL I
1SG.S=IRR-be.embarrassed-IRR
‘I am not embarrassed.’
3.39
Te antaroite aye.
te
aNtaro-ite ay-e
NEG.REAL big-AUG take-IRR
‘She didn’t gather a lot [of fruit].’
The particle te is classed as such because in most cases it functions as a ‘bound’ form,
lacking stress and pause phenomena. However, since in some cases it meets the criteria
for an independent phonological word, it occupies the position between a clitic and an
independent word.
3.3.4 Other morphological processes in Ashéninka Perené
After addressing affixation and cliticization in §3.3.2 and §3.3.3, I will discuss in this
section other morphological operations such as infixation, circumfixation, reduplication,
noun compounding, and noun incorporation.
Infixation is infrequently found with numeral and adverbial roots, as illustrated in
(3.40-3.41). In (3.40), the reduced connector ahatzi~aisatzi ‘also’, ‘and’ is infixed to the
numeral aparoni ‘one’; the nominal augmentative suffix -ite is inserted in the adverb
maaroni ‘all’ in (3.41).
116
3.40
aparohatzini
aparo<hatzi>ni
one<also>one
‘one day’
3.41
maaroeteni
maaro<ite>ni
all<AUG>all
‘everyone’
There are a few circumfixes in Ashéninka Perené. By definition, circumfixes occur on
both sides of the base, which is the element to which a morphological operation applies
(Haspelmath 2002:19). The discontinuous irrealis status marker N-…-e/-ia and the
discontinuous plural marker he-…-ni are treated here as circumfixation. Another
discontinuous plural marker –aiy-…-ni, mentioned in Kampanist literature (e.g. Payne
1989:200-201), has been marginally attested in my corpus. Examples are provided in
(3.42-3.44).
3.42
Anyaatsahetani.
a=nyaatsa-he-t-a-ni
1PL.S=play-PL-EP-REAL-PL
‘We played together.’
3.43
Ishitovaiyaheni.
i=shitov-aiy-ah-i-ni
3m.S=escape-PL-REGR-REAL-PL
‘They went back.’
3.44
Te onkimisantena.
te
o=N-kimisaNt-e=na
NEG.REAL 3n.m.A=IRR-pay attention-IRR=1SG.O
‘She didn’t listen to me.’
Compounding in Ashéninka Perené is fairly productive and is typically expressed in
the possessor-possessed or whole-part relationships. Noun compounds are defined (i) as
117
two lexemes juxtaposed in a single word-form; (ii) the compounds are typically
idiomatic but may express a possessor-possessed relationship; (iii) a dependent noun in a
compound is not referential but generic (this is shown for the verb-headed compounds or
noun incorporation); (iv) compounds exhibit a certain degree of phonological,
morphological, or syntactic cohesion, e.g. a special compound stress on the first element
of the compound, etc. (Haspelmath 2002:155-161). An example of an N-N compound is
provided in (3.45).
3.45
Nontaavakirota pisaarikontate piitzi.
no=N-ta-av-ak-e=ro=ta
pi=saari-koNta-te
1SG.A=IRR-burn-DIR-PRF-IRR=3n.m.O=OPT 2poss=opossum-sole-poss
pi=itzi
2poss=foot
‘I’ll burn your soles of an opossum, your stinky feet.’
In (3.45), the second noun in the compound is a syntactically bound noun encoding a part
of the whole while the first noun is a generic noun denoting an entity. The semantics of
such compounds sometimes may go beyond the sum of their parts, as (3.45)
demonstrates. The reading of the compound in (3.45) is based on the speaker’s
associations of the opossum with the stench. The stress in the Ashéninka noun
compounds is quite predictable due to the constant second-syllable stress assignment.
This assignment pattern stresses even-numbered syllables counting from the beginning of
the word, with the primary stress augmenting the stress closest to the end of the word
(word-final syllable is never stressed), e.g. pi. ''saa.ri. 'kon.ta.te. The second element of
the compounds is typically a phonologically unreduced nominal stem which can be used
in discourse independently.
118
Many Ashéninka noun compounds are plant-based, encoding the possessor-possessed
or whole-part relationships. As seen in Table 13, the first noun is a generic noun denoting
a plant e.g. incha ‘plant’, while the second noun (or its phonologically reduced
equivalent) is the plant or body part. The set of plant-part terms performs a classifying
function by identifying a certain type of entity. The lexical origin of the plant terms is
typically transparent, e.g. -paNki is derived from the verb paNki ‘to grow’. The
compounds express a subcategorizing ‘X is a type of Y’ relationship where the first
element of the compound is a variable X and the second element is a stable morpheme Y
of a clear lexical origin. Some of the plant-based parts in Table 13 have lost
phonological substance, becoming monosyllabic, e.g. -shi < oishi ‘leaf’.
TABLE 13. Plant-based parts in compounds
Form
-paNki-toNki
-tyaki
-shi
-kota
-ponkitzi
Origin
‘plant’ from panki ‘to grow’
‘long tree segment’ from tonki ‘bone’
‘flower’ from tya ‘to blossom’
‘leaf’ from oishi ‘leaf’
‘piece’ from kota ‘piece’
‘ground-level part of tree trunk’ from
ponkitzi ‘ground-level part of tree trunk’
Examples
incha-panki ‘plank’
incha-tonki ‘beam’
incha-tyaki ‘tree flower’
incha-shi
‘plant leaf’
incha-kota ‘split wood’
incha-ponkitzi ‘tree trunk’
Most plant parts are marginally productive. The collected data show that the speakers
do not use them to describe new cultural items. The majority of plant-based compounds
appear to be lexicalized. For example, in (3.46-3.47) the compound incha-kota ‘split
wood’ appears to be semantically non-compositional, denoting the concept of splitting.
3.46
Pinchakotaero piitzira.
p=inchakota-e=ro
pi=itzi=ra
2A=split-IRR=3n.m.O 2poss=foot=DEM
‘Flatten your foot.’
119
3.47
Pinchakotaena.
p=inchakota-e=na
2A=split-IRR=1SG.O
‘Split wood for me.’
Noun incorporation is infrequent in Ashéninka Perené, at least as based on the
collected corpus. Noun incorporation refers to morphological structures in which a
nominal constituent is added to the verb root (Aikhenvald 2007:11). In Ashéninka
Perené, a bare noun root is incorporated. In (3.48), vatsatsi ‘meat, flesh’ is incorporated
without the unpossessed noun suffix –tsi.
3.48
Itovatsatziro ivori.
i=tov-vatsa-tz-i=ro
3m.A=cut-flesh-EP-REAL=3n.m.O
‘He did flesh-cutting from his thigh.’
i=pori
3m.poss=thigh
In some cases, the incorporated second noun may be an embedded compound, as seen
in (3.49). The nominal bare stems -vatsa ‘meat’(unpossessed form vatsatsi) and -pori‘foot, muscle’(unpossessed form poritsi) are combined into the noun compound -vatsavori- ‘meat-muscle’, which follows the verbal stem shiriNka ‘grow’. The verb-noun
compound literally means ‘he did meat-muscle-growing’, the embedded noun-noun
compound denoting a generic entity.
3.49
Ishirinkavatsavoritanahi.
i=shiriNka-vatsa-pori-t-an-ah-i
3m.S=grow-flesh-muscle-EP-DIR-REGR-REAL
‘He grew back muscle in his flesh.’
Reduplication is limited in Ashéninka Perené. It is attested only with verbs. This
morphological process is iconic in a sense that it is showing a correlation between a
120
derivational process and its semantics. The semantic effect of reduplication is such that
the reduplicated verbal roots have the meaning of intensity or repetetiveness of the
described acitivity. The reduplication pattern varies depending on the phonological nature
of the base. Typically, the entire root is copied. When the verb root is V-final, e.g. ira
‘cry’, the reduplicant matches the base, as shown in (3.50). When the verb root is C-final,
e.g. ky –carry a load, piy ‘return’, an epenthetic element –a is added to the base, as seen
in (3.51-3.52).
3.50
Noina irairatanaka.
no=ina
iraira~t-an-ak-a
1SG.poss=wife 3n.m.S.cry~EP-DIR-PRF-REAL
‘My wife cried a lot.’
3.51
Ikyakyatakina.
i=kyakya~t-ak-i=na
3m.A=carry.load~EP-PRF-REAL=1SG.O
‘It took him a lot of effort to take me down.’
3.52
Airo pipiyapiyatashitari.
airo
pi=piyapiya~t-ashi-t-a=ri
NEG.IRR 2A=come.back~EP-APPL.INT-EP-REAL=3m.O
‘You won’t go back all the time to this place [to hunt].’
When the C-initial base ends in a nasal element, e.g. ken~kin ‘walk’, the reduplicant
copies the base, but the base itself becomes light, without the root-final nasal, as seen in
(3.53). When the V-initial base ends in a nasal element, e.g. amin ‘look for’ the
reduplicant and the base match, as exemplified by (3.54).
3.53
Nokekenatakairi tsiteni.
no=keken~a-t-aka-i=ri
tsiteni
1SG.A=walk~REP-EP-CAUS.COM-REAL=3m.O at.night
‘I walked a lot with the [crying] baby that night.’
121
3.54
Paminaminataiterira.
p=aminamina~t-a-it-e=ri=ra
2A=look.for~EP-REP-EP-ICPL-IRR=3m.O=ADV
‘You will look and look for him, right?’
3.4 Word classes
The major word classes in Ashéninka Perené are verbs and nouns. There are also small
classes of adjectives and adverbs, in addition to other word classes such as personal and
possessive pronouns, demonstratives, interrogative pronouns, numerals, connectives,
interjections, discourse particles, and ideophones. The word classes are distinguished on
the basis of syntactic, semantic, and morphological criteria.
3.4.1 Nouns
The semantic criterion states that nouns are “time-stable concepts which do not vary
across time” (Payne 1997:33). As such, Asheninka Perené nouns refer to people, things,
elements of the vegetal and animal worlds, nature, e.g. mapi ‘stone’, impereta ‘cliff’,
nihaa ‘water’, pareni(ni) ‘river’. Morphosyntactic properties of nouns in Ashéninka
Perené include distributional and structural properties. Distributional properties have to
do with how nouns are distributed in phrases, clauses, and texts (Payne 1997: 33). In
Ashéninka Perené, nouns serve as heads of phrases, subjects and objects of clauses, and
topics of texts. Nouns can be modified by adjectives, e.g. aNtaro korarintsi ‘big corral’;
numerals, e.g. mava kooya ‘three women’; and demonstratives, e.g. iyoka apiteroite
kishiisatzi ‘these two pajonalinos’. Structural properties have to do with the internal
structure of nouns (Payne 1997: 33). They involve the categories of case (locative case
122
–ki), morphologically expressed gender (masculine –ri, non-masculine -ro), optionally
expressed plural number -paye, and possession. The locative case marking –ki on nouns
has a diffuse spatial meaning of ‘in/ at/ on/ by/ to.’ When nouns lack gender markers,
gender is revealed through the cross-referencing clitics on the verbs, gender markers on
adjectives, demonstratives, possessive pronouns, and locative-existential adverbs, as
shown in Table 13.18
TABLE 14. Gender agreement patterns
Gender Adjective Demonstrative Personal
pronoun
Masc
-ri;-nzi
(i)ri-/(i)yoiriNon-m -ro;-Nto
iroiro-
Possessive
pronoun
irØ-
Predicate
Adverb
i-/y-; -ri
o-; -ro
hirihiro-
Possession is morphologically expressed on two groups of nouns: obligatorily
(inalienably) possessed and optionally (alienably) possessed. Morphologically expressed
possession covers the following semantic relations.
(i) Kinship relationships, e.g. no=niro [1SG.poss=mother] ‘my mother’, iri=niro Edwin
[3m.poss=mother Edwin] ‘Edwin’s mother’.
(ii) Whole-part relationships (a body part of a human, animal or part of an object), e.g.
n=ako [1SG.poss=hand] ‘my hand’, o=yeNka tyaapa [3n.m.poss=fat chicken] ‘chicken
fat’, inchashi o=paritsa [herb 3n.m.poss=root] ‘herb roots’.
(iii) Ownership relationship including temporary possession, e.g. pi=shima-te
[2poss=fish-poss] ‘your fish’, i=vanko-shi koventsari-paye [3m.poss=houseCL:small.thin.flexible hunter-PL] ‘hunters’ shack’.
18
For details on gender agreement in Ashéninka Perené, see Mihas 2010c.
123
(iv) Association relationship, e.g. no=vairo ‘my name’, no=tsipataro ‘my companion’,
no=shaniNka ‘my people’. Association relationship can also be expressed by
juxtaposition, e.g. heva-t-ak-aNt-tz-i-ro=ri kampana [lead-EP-CAUS.SOC-APPL.REAS-EPREAL=3n.m.O=REL
church] ‘the leader of the church’. Note that when possession is
expressed via juxtaposition, the possessor NP precedes the possessed NP.
As the above examples show, the morphological process of possession applies to the
possessed NP whereas the possessor NP remains unmarked. Obligatorily possessed nouns
are marked by a possessive proclitic whereas optionally possessed nouns are marked by a
possessive proclitic and one of the possessive suffixes -te, -ni, -ri, so the grammatical
marking for alienable possession is longer than that for inalienable possession19.
Examples of the obligatorily possessed nouns are body parts, e.g. no=ito ‘my head’;
kinship terms, e.g. no=saro ‘my grandma/granddaughter’; essential personal belongings
such as no=vito ‘my canoe’, no=vaNko ‘my house’. Examples of optionally possessed
nouns are no=chakopi-te ‘my arrow’, no=yoka-ni ‘my coca’, n=otsitzi-te ‘my dog’,
n=akipa-ri ‘my grilled fish’. In general, when alienably possessed nouns appear in an
‘unpossessed’ form (including, as an exception, inalienably possessed body parts), they
take the ‘absolute’ or unspecified possessor suffixes –(n)tsi or –rintsi/-rontsi, e.g. tsitsi
‘fire wood’, avotsi ‘trail, road’, aikintsi ‘tooth’, poritsi ‘thigh’, maamiNtotsi ‘bed’,
kitsarentsi ‘clothes’, aaviNtarontsi ‘medicine’. Examples of nouns denoting body parts
with the unspecified possessor marking –(n)tsi are provided in (3.55-3.56).
19
Payne states that –ni is assigned to a mono- or bisyllabic nominal host, while -te occurs with
multisyllabic nominal roots (1989:67). This rule holds for Ashéninka Perené; cf. notsitzite ‘my dog’ (otsitzi
‘dog’), noyokani ‘my coca’(koka ‘coke plant’).
124
3.55
Paminaite katsarikentsicharini.
p=amin-a-it-e
katsariki-ntsi-chaari-ni
2S=look.for-EP-ICPL-IRR testicle-ABS-dark.water-AUG
‘Go look for the River of Testicles.’
3.56
Airo pikotsitziri vatsatsira.
airo
pi=kotsi-tz-i=ro
vatsa-tsi=ra
NEG.IRR 2A=cook-EP-REAL=3n.m.O meat-ABS=DEM
‘Don’t cook that meat.’
There are also non-possessable nouns such as names, astronomical bodies, rationals20,
and natural elements which appear bare in the clause, without any inflectional
morphology, e.g. kashiri ‘moon’, pava ‘solar deity’, kamaari ‘devil’, tampyaa ‘wind’.
Some degree of morphological irregularity is attested in a few nouns. A syllable may
be deleted in a possessed form of the noun root, e.g. tsarato ‘bag’, cf. no=tsara-te ‘my
bag’. Conversely, a syllable may be added to the possessed form, e.g. tsitsi ‘fire wood’,
cf. no=tsima ‘my fire wood’. Some optionally possessed nouns, which appear bare in an
upossessed form, lack a possessive suffix in a possessed form, e.g. kaniri ‘manioc’, cf.
no=yaniri ‘my manioc’; kishiri ‘comb’, cf. no=ishiri ‘my comb’; tsivo ‘little box’, cf.
o=tsivo ‘her box’. Some nouns retain the absolute/unspecified possessor suffix –ntsi
when occurring in a possessed form, e.g. kamaantsi ‘phlegm’; cf. Pishitero piyamaantsite
pi=shit-e=ro pi=yamaantsi-te [2S=wipe.out-IRR=3n.m.O 2poss=phlegm-poss].
Although the plural marking of subjects and objects is not mandatory (on both nouns
and verbs), it is generally sensitive to definiteness and animacy of noun referents.
Examples are given in (3.57-3.59).
20
Corbett defines ‘rationals’ as ‘various mythical beings’ and notes that this term is ‘almost equivalent to
humans’ (2007:259).
125
3.57
Atziripaye ironyaaka kimoshiri ikantahetave.
atziri-paye ironyaaka kimoshi-ri
i=kaNt-a-he-t-a=ve
person-PL now
be.joyous-ADJ 3m.S=be-EP-PL-EP-REAL=EXCL
‘The people were very happy.’
3.58
Pivitsikanahinaro novanko novantayetyarori notsanoirikipaye.
pi=vitsik-an-ah-e=na=ro
no=paNko
2A=build-DIR-REGR-IRR=1SG.O=3n.m.O 1SG.poss=house
no=v-aNt-a-ye-t-ia=ro=ri
1SG.A=put-APPL.REAS-EP-DIST-EP-IRR=3n.m.O=REL
no=tsanoiriki-paye
1SG.poss=adornment-PL
‘Build me a gazebo to display my handicrafts.’
3.59
Nokovatzi nonpankitemi mapocha.
no=kov-atz-i
no=N-paNki-t-e=mi
mapocha
1SG.S=want-PROG-REAL 1SG.S=IRR-plant-EP-IRR=CNT.F papaya
‘Originally, I wanted to plant papaya.’
The diffuse spatial meaning of the locative marker –ki may be extended to the
temporal sense ‘during’, as shown in (3.60).
3.60
Irotaki novavisantarori nanyantari shimaanakipaye, kashirikipaye,
kitaiterikipaye.
irotaki no=v-avis-aNt-a=ro=ri
FOC 1SG.A=CAUS-pass-APPL.REAS-REAL=3n.m.O=REL
n=any-aNt-a-ri
shimaana-ki-paye kashiri-ki-paye
1SG.S=live-APPL.REAS-REAL-NMZ week-LOC-PL moon-LOC-PL
kitaiteri-ki-paye
day-LOC-PL
‘With this I live my life over weeks, months, and days.’
Other morphemes that occur with nouns include distributive –ye, augmentatives –ite,
-mashi, -saNte, diminutives –patsaini, -(i)ni, -aniki,-peta, a sizeable number of
derivational classifying suffixes, demonstrative enclitics =ka, =ra, =Nta, adverbial past
126
enclitic =raNki, and pragmatic enclitics =ha, =tya, =kya, =ve. Some nouns are derived
from verbs with the help of nominalizing suffixes –meNto, -Nka, -ri, e.g. i=tot-a-meNto
[3m.poss=cut-EP-NMZ] ‘his axe’, i=tsaro-Nka-ki-ni [3m.poss=to.fear-NMZ-LOC-DIM]
‘in fear ’, pi=N-kats-a-ri [2S=IRR-feel.pain-EP-NMZ] ‘chief’.
Nouns can function as predicates when taking verbal morphology, as seen in (3.613.62).
3.61
Pairani noshaninkapaye ivavanitari mapi.
pairani no=shaniNka-paye
i=pava-ni-t-a=ri
mapi
before 1SG.poss=fellowman-PL 3m.poss=deity-poss-EP-REAL=3m.O stone
‘Long ago my fellowmen idolized [a piece of] stone.’
3.62
Ampotsotyaro potsotzi.
a=N-potsot-ia=ro
potsotzi
1PL.A=IRR-annatto-IRR=3n.m.O
annatto
‘We’ll rub annatto [into the wound].’
However, certain restrictions apply to the nouns which appear in a predicate slot.
Noun-derived predicates exhibit a limited use of verbal derivational morphology; e.g.
they don’t take valence-adjusting operators, degree, time and manner suffixes such as –
niNt ‘a little bit’, –aman ‘early’, -apaiNt ‘once’, -apanaNt ‘one more time, again’.
Neither do nouns as predicate heads occur in the imperative mood.
3.4.2 Verbs
Verb roots serve as heads of predicates. They typically refer to least stable concepts such
as actions and processes. In terms of their structure, verb roots are always bound and
minimally have a subject proclitic and a status suffix attached. Verbs can also appear
with multiple suffixes, enclitics, and prefixes and proclitics (see §3.4.2 and §3.4.3
127
below). Verbs can be marked for number (plural suffix –he, discontinuous plural marker
he-/aiy-..-ni, distributive -ye), person (cross-referencing S/A proclitics and O enclitics),
gender (3p masculine clitics i=, y=, ir=, =ri; 3p non-masculine clitics o=, =ro), aspect
(perfective –ak, durative –vai, progressive –aty~-atz, regressive–ah~-a, incompletive
–it21, habitual –api(i)Nt, customary –aNt, inchoative –ima, repetitive -a), status (realis -i,
-a, irrealis –e, -ia), and mood (interrogative =ka, optative =ta, apprehensive =kari,
dubitative =ma, conditional =rika, counterfactual =mi). Examples of inflectional verbal
morphology are provided in (3.63-3.65).
3.63
Nokovi nonkamarankimi.
no=kov-i
no=N-kamaraNk-e=mi
1SG.S=want-REAL 1SG.S=IRR-vomit-IRR=CNT.F
‘I wanted to vomit [but I didn’t].’
3.64
Yayimikari maini.
y=ay-e=mi=kari
maini
3m.A=take-IRR=2O=APPR
bear
‘Beware of the bear, he may abduct you.’
3.65
Tsikama opaitatyaka?
tsika=ma
o=pai-t-a=tya=ka
WH=DUB 3n.m.S=be.called-EP-REAL=EMPH=Q
‘What could it possibly be?’
The valence of Ashéninka Perené verbs can be changed with valence-adjusting
derivational prefixes and suffixes. By definition, syntactic valence refers to the actual
21
This morpheme indicates that the action has not been completed either because it hasn’t yet started, or
because it failed to start, or ended unfinished. In Kampanist scholarsip, a discussion of this suffix is found
in Payne, Payne, & Santos (1982) who point out that Ashéninka Apurucayali has the partitive suffix -itsi
indicating that the action was executed partially or in parts (61); in García Salazar (1997) who identified
this suffix as the partitive –iti~-it~-t in Ashéninka Ucayali (32); in Swift (2008) who proposes to call the
Caquinte suffix -γiha ‘limitative’ due to its meaning of the action, process, state being placed within some
limits (84).
128
number of arguments present in a clause. The semantic effect of increasing syntactic
valence is generally characterized by upgrading a peripheral participant to direct-object
function or by adding a new participant; in contrast, the effect of decreasing valence is to
demote the central participant to a more peripheral function or eliminate it altogether
(Payne 1997:170, 172). In (3.66-3.70), valence-increasing morphemes are comprised of a
few causative prefixes such as (1) the malefactive causative mi-, which involves the
causer’s detrimental action towards the (human or animate) causee, performed against the
causee’s will; (2) the non-productive22 agentive causative prefix o-~oi~ov- whose
function is to indicate the causer’s agentivity (the causer may be non-human or
inanimate) in an action performed on the causee who is characterized by the low retention
of control over the situation; (3) the productive sociative causative suffix –ak, typically
used in a situation when both the causer and the causee cooperate in an action which has
an enabling or beneficial effect on the causee.
3.66
Ari imisaikiri.
ari
i=mi-saik-i=ri
PP
3m.A=M.CAUS-sit-REAL=3m.O
‘There they had him seated [against his will].’
3.67
Noitakeri pashini inchashi.
n=oi-t-ak-e=ri
pashini incha-shi
1SG.A=CAUS.AGT.drink-EP-PRF-IRR=3m.O other plant-leaf
‘I’ll make him [the sick person] drink another herb concoction.’
22
Payne observes that the Ashéninka set of derivational causative prefixes omin-, omi-, oi-, -o-, ow- is not
productive (2002: 488). In the collected corpus, the prefix has been attested with a handful of verbs such as
ma ‘sleep’, cf. oima ‘make [a baby] sleep’; ha ‘go’, cf. oiha ‘chase away, pursue’; ir ‘drink’, cf. oit ‘make
drink’; avis ‘pass’, cf. vavis ‘make pass [recover from an illness]; spend [days]’; anii ‘walk (down)’, cf.
vanii ‘make walk down’; kimoshiri ‘be happy, joyous’, cf. oimoshirink ‘rejoice, celebrate [make happy]’.
129
3.68
Irotaki ovavisakotakiriri.
irotaki o=v-avis-ako-t-ak-i=ri=ri
FOC 3n.m.S=CAUS.AGT-pass-APPL.REAS-EP-PRF-REAL=3m.O=REL
‘This is what [the herbs] made him [the sick person] recover from the illness.’
3.69
Ivaniitapakina.
i=v-anii-t-ap-ak-i=na
3m.A=CAUS.AGT-walk-EP-DIR-PRF-REAL=1SG.O
‘He took me down [the son helped the wounded father get off the tree because the
father was injured and unable to climb down on his own].’
3.70
Yantavaitakakiri.
y=aNt-a-vai-t-ak-ak-i=ri
3m.A=work-EP-DUR-EP-CAUS.SOC-PRF-REAL=3m.O
‘He made them work [enabling the workers to get food].’
In addition to the valence-increasing causative forms discussed above, applicatives
also promote a peripheral participant to direct object function. Examples of the valenceincreasing applicatives such as generalized –ako, presential –imo, separative -pitsa,
instrumental –aNt, benefactive –veNt~~viNt are provided in (3.71-3.75).
3.71
Tzimatsi isorarotepaye katziyaventariri.
tzimatsi i=soraro-te-paye
katziy-a-veNt-a=ri=ri
EXIST 3m.poss=soldier-poss-PL stand.guard-EP-APPL.BEN-REAL=3m.O=REL
‘He had soldiers who were standing guard for him.’
3.72
Te intyerovimotapakyari.
te
i=N-tyerov-imo-t-ap-ak-e=ri
NEG.REAL 3m.A=IRR-kneel.down-APPL.PRES-EP-DIR-PRF-IRR=3m.O
‘He didn’t kneel in his presence.’
3.73
Ikantakoitziri aparoni koshiri.
i=kaNt-ako-it-tz-i=ri
aparoni koshiri
3m.A=say-APPL-ICPL-EP-REAL=3m.O one
monkey
‘They talked about a monkey.’
3.74
Nomanapitsatatyeri.
no=man-a-pitsa-t-aty-e=ri
1Sg.A=hide-EP-APPL.SEP-EP-PROG-IRR=3m.O
‘I will hide from him.’
130
3.75
Mantsiyarira impoki pashini, pashini nosaatantariri.
mantsiya-ri=ra
i=N-pok-e
pashini pashini
be.sick-ADJ=DEM 3m.S=IRR-come-IRR another another
no=saa-t-aNt-a=ri=ri
1SG.A=bathe-EP-APPL.INST-REAL=3m.O=REL
‘When another sick person comes, I steam-bathe him with a different herb.’
Valence-decreasing operations include the customary –aNt23 and reciprocal –av. The
customary –aNt, when denoting a habitual action, reduces the syntactic valence of
transitive verbs. Examples are given in (3.76-3.77).
3.76
Notsitzite te yatsikante.
n=otsitzi-te
te
y=atsik-aNt-e
1SG.poss=dog-poss NEG.REAL 3m.S=bite-CUST-IRR
‘My dog doesn’t bite.’
3.77
Inintatavakatya.
i=niNtat-av-ak-aty-a
3m.S=kiss-RCP-PRF-PROG-REAL
‘They are kissing.’
Another mechanism for decreasing verb valence includes the cumulative formatives
–a and -ia which, with some verbs, signal that an actor is acting upon himself. The
suffixes –a and –ia are category-based allomorphs. When a verb is inflected for the
reality status category, the realis status is marked by –a and irrealis status by –ia. By
definition, the cumulative formative expresses more than one category which are grouped
together in a single, indivisible unit (Bickel & Nichols 2007:188). In the Ashéninka
Perené case, the cumulating values of realis status and reflexivity can be attributed to the
23
The customary suffix –aNt , identical in form and similar in meaning to that found in Ashéninka Perené,
is attested in other Kampan languages such as Caquinte -aNt ‘detransitivizer’ (Swift 1988:73), Ashéninka
Pichis –aNt ‘antipassive or indefinite O marker’ (Payne 2002: 492; Payne &Payne 2005:43); Nanti –aNt
‘characteristic’ (Michael 2008:277).
131
aforementioned formatives –a and -ia. For example, in (3.79-3.), when the verb root tot
‘cut’ has a reflexive reading, it takes the cumulative formative -a with the
realis/reflexivity exponence, whereas in a non-reflexive situation, the suffix –i occurs,
which has realis/non-reflexivity values.
3.78
Iriro itotaka.
iriro i=tot-ak-a
he 3m.S=cut-PRF-REAL.RFL
‘He cut himself.’
3.79
Itotakiniri iishi itomi.
i=tot-ak-i=ni=ri
i=ishi
3m.A=cut-PRF-REAL=3O=3m.O 3m.poss=hair
‘He cut his son’s hair.’
i=tomi
3m.poss=son
This distinction has served as a basis for distinguishing between reflexive and nonreflexive classes of Kampan verbs in literature (e.g. Payne, Payne, & Sanchez 1982: 4445; Payne 1989:105-110; Garcia Salazar 1997:37-38; Swift 2008:55-57; Wise 1986:586).
Table 18, adapted from Payne (1989:110), is a summary of morphological reflexives in
Ashéninka Perené. As Table 15 shows, morphological reflexives are fused with the
reality status category.
TABLE 15. A summary of the cumulating reality status/reflexivity values of verbs
Reality status
Realis
Irrealis
Reflexive
-a
-ia~-eya
Non-reflexive
-i
-e
Prototypical reflexive verbs use the cumulative formatives –a and -ia as
morphological reflexives. A prototypical reflexive situation is such when “the
grammatical subject is being acted upon in some way, i.e. is himself an undergoer of the
132
event” (Epps 2008:473). In Ashéninka Perené, prototypical reflexive situations involve
verbs with a standard reflexive reading, e.g. noNkitsatya ‘I’ll dress myself’, nokaata ‘I
bathe myself’, nototaka ‘I cut myself’, nonyaka ‘I saw myself’, nopashiventashita ‘I am
ashamed of myself’. These verbs become monovalent when marked by –a or -ia.
In addition, the cumulative formatives –a and -ia exhibit other functions such as a
passive reading (when some other participant acts on the subject), e.g. neve ‘be pleased
with’, pai ‘be called’, amit ‘be accustomed to’, pi ‘be transformed into’, pina ‘be paid’,
or reciprocal interpretations (when subjects act upon each other), e.g. nyaatsa ‘play with’,
tsipa ‘accompany somebody’, vitsa ‘greet [each other]’, niNtat ‘kiss’, avitsanot ‘hug’,
aree~arii ‘visit with somebody.’ Semantics of other verbs, marked by the suffixes –a and
-ia, is only marginally related to the affectedness of subject, which is the main property
of reflexive situations. Such verbs include konyaa ‘appear’, kimi ‘seem’, pony ‘come
from’, shiroNt ‘smile’, ashi ‘possess’, kaNt ‘happen’, ‘be’. Miscellaneous other verbs
from this class do not have clear reflexive interpretations either, e.g. piy ‘return’, vir
‘come near’, pavani ‘idolize’, yomit ‘teach’, kamaNt ‘inform’, any ‘live’, etc. It appears
that, in some cases, the occurrence of the suffixes –a and -ia is grammatically
conditioned. My data show that the suffixes –a and -ia co-occur with the applicative
reason/instrumental –aNt, frustrative –ve, and reciprocal –av.
In general, the valence-decreasing function of the cumulative formatives –a and -ia is
largely limited to the standard reflexive verbs. The valence of other verbs from the class,
which takes the suffixes –a and -ia, is typically not affected, as seen in (3.80-3.82).
133
3.80
Isaiki itsipataro iina.
i=saik-i
i=tsipat-a=ro
3m.S=be.at-REAL 3m.A=accompany-REAL=3n.m.O
‘He lived with his wife.’
i=ina
3m.poss=wife
3.81
Oyomitanari nonironi.
o=yomit-a=na=ri
no=niro-ni
3n.m.A=teach-REAL=1SG.O=REL 1SG.poss=mother-deceased
‘What my deceased mother taught me.’
3.82
Irotaki nokamantantamiri.
irotaki no=kamaNt-aNt-a=mi=ri
FOC 1SG.A=inform-APPL.REAS-REAL=2O=REL
‘That’s why I advised you.’
In light of this synchronically weak relation to reflexivity, a more useful approach to
the classification of inflectional verbal morphology in Ashéninka Perené may be to
abandon the reflexive interpretation of the suffixes –a and -ia. This alternative proposal
has been advocated by a few scholars, e.g. by Michael (2008:250) and Cysouw
(2007:143). A similar classification will work as well in the case of Ashéninka Perené.
Two verb classes may be distinguished, based on the synchronic shape of the inflectional
forms the verbs take: class A and class I, as illustrated in Table 16. When the verb is
inflected for the same category such as realis status (or irrealis status), it exhibits
lexically-conditioned allomorphy, i.e. the choice of the realis status formative –a or –i is
a lexical and essentially unpredictable matter.
TABLE 16. Lexical classes of verbs
Reality status
realis
irrealis
Class A
-a
-ia
Class I
-i
-e
134
There are also valence-preserving derivational suffixes: frustrative –(mache)ve which
expresses a sense of some unrealized action when the actor’s desires or intentions cannot
be satisfied due to some preceding action, event, or circumstance; reversative –ry which
signals the reversal of an action or state, expressed by the verb root; and impersonal third
person plural passive –ai which, along with the subject proclitic, is used to refer to an
unspecified group of actors. Examples are provided in (3.83-3.86).
3.83
Nocharini shintsiri inavita.
no=charini
shintsi-ri
i=na-vi-t-a
1SG.poss=grandpa be.strong-ADJ 3m.S=be-FRUS-EP-REAL
‘My grandpa used to be strong.’
3.84
Nooyamachevetaro omparyanake.
n=ooya-macheve-t-a=ro
o=N-pary-an-ak-e
1SG.A=expect-FRUS-EP-REAL=3n.m.O 3n.m.S=IRR-fall-DIR-PRF-IRR
‘I expected it to fall in vain.’
3.85
Itsirikaryakoitahinaro.
i=tsirik-a-ry-ako-it-ah-i=na=ro
3m.A=attach-EP-REV-APPL-ICPL-REGR-REAL=1SG.O=3n.m.O
‘They removed some of the bandage [from my arm].’
3.86
Ityankaitakina Irimashiki.
i=tyaNk-ai-t-ak-i=na
Irimashi-ki
3m.A=send-IMP.P-EP-PRF-REAL=1SG.O Lima-LOC
‘They sent me to Lima.’
In addition, verb roots can incorporate nouns; take classifiers –ki ‘small, round’,
-poroki ‘quantity’, -ako ‘vessel’, -shi ‘small, thin, flexible’; directionals –ap (goal), –an
(source), -av (O-oriented goal); scalar (degree) suffixes –niNt ‘almost, a little bit’,
-pero~-piro ‘indeed’; adverbial suffixes of time –aman ‘early’, -it ‘before’, and manner
135
–ite ‘quickly’; subordination clitics =ra and =ri, pragmatic clitics =kya, =tya, =ha, =ve,
adverbial clitics of locality =ka, =ra, =Nta and time =raNki, among others (see Tables 9,
15 for details).
Ashéninka Perené verbs divide into transitivity classes depending on the transitivity
type of clauses they may appear in. Two basic transitivity clause types include
intransitive clauses which have one obligatory argument in S (intransitive subject)
function and transitive clauses which have two arguments in A (transitive subject) and O
(transitive object) functions (Dixon 2010a:115-6). The core arguments may be expressed
by the person clitics, personal pronouns, demonstratives, or noun referents. Examples of
verb roots expressing intransitivity values are kam ‘die’, shiy ‘run’; kamaraNk ‘vomit’;
transitive verb roots are am ‘bring’, a ‘take’, pos ‘hit’. Verb roots can be ambitransitive,
i.e. a verb root can have either transitivity value. Transitive verb roots frequently appear
in intransitive clauses; vice versa, intransitive roots occur in transitive clauses. In general,
verb roots by themselves do not have fixed transitivity, when each is either strictly
transitive or strictly intransitive. The lexical meaning of the verb root determines if it has
low or high transitivity values. Examples of the ambitransitive verb roots (A=S) a ‘take’,
‘harvest’ and aNt ‘make’, ‘work’, are provided in (3.87-3.90).
3.87
Paanakinaranki.
p=a-an-ak-i=na=raNki
2A=take-DIR-PRF-REAL=1SG.O=ADV
‘You took me along the other day.’
3.88
Naake ironyaaka.
n=a-ak-e
ironyaaka
1SG.S=take-PRF-IRR now
‘I’ll harvest now.’
136
3.89
Pantahinaro novanko.
p=aNt-ah-e=na=ro
no=paNko
2A=make-REGR-IRR=1SG.O=3n.m.O 1SG.poss=house
‘You will make the house for me.’
3.90
Nantavaitaki osheki.
n=aNt-a-vai-t-ak-i
osheki
1SG.S=work-EP-DUR-EP-PRF-REAL many
‘I worked a lot.’
Some transitive verb roots have culture-specific semantic roles mapped on the second
core argument O such as tomiNt ‘to be son of’ or matzi ‘to cast spell on’, as shown in
(3.91-3.92).
3.91
Iyoka te intomintyari atziri.
yoka te
i=N-tomiNt-ia=ri
atziri
this.m NEG.REAL 3m.A=IRR-be.son.of-IRR=3m.O person
‘This is not the man’s son.’
3.92
Irotaki matzitakiriri.
irotaki matzi-t-ak-i=ri=ri
FOC cast.spell-EP-PRF-REAL=3m.O=REL
‘She is the one who cast spell on him.’
Copular verbs. In addition to regular verbs, which take freely inflectional and
derivational morphology, there is a class of irregular existential and light copula verbs in
Ashéninka Perené, limited in their ability to take verbal morphology. Existential verbs
include two existential/possessive verbs tzimatsi and ainiro, and the existential/possessive
negation verb tekatsi, as shown in Table 17.
137
TABLE 17. Existential/possessive verbs
Existential verbs
Affirmative
Negation
Form
tzimatsi
ainiro
kaNt
tekatsi
Function
indicate existence/possession
indicate existence/possession
indicate existence
deny existence/possession
The verbs in Table 17, except the polyfunctional verb root kaNt ‘happen, can, be, say’,
have restricted morphological possibilities in that they do not take S personal proclitics.
Some are used as invariant forms whose origin may not be accessible, e.g. ainiro ‘exist’
is not connected to its original verbal roots. The copula verbs, whose origin is
transparent, tend to take minimally aspect and/or status formatives, e.g. tek-atsi [NEGSTAT], tzim-atsi [have-STAT], o=kaNt-a [3n.m.S=be-REAL]. When the affirmative
existential verbs tzimatsi and ainiro and the negation verb tekatsi have a possessive
reading, they appear with one O argument, which is a clausal complement, as shown in
(3.93-3.94).
3.93
-Noshinto, tzimatsi pinintani? -Tekatsi, apa.
no=shiNto
tzimatsi pi=niNt-a-ni
tekatsi
apa
1SG.poss=daughter EXIST 2S=want-REAL-AUG NEG.EXIST father
‘Daughter, do you have a boyfriend?’ ‘I don’t have [a boyfriend], father.’
3.94
Ainiroma notashatzi.
ainiro=ma
no=tash-atz-i
EXIST=DUB 1SG.S=be.hungry-PROG-REAL
‘I have hunger.’
When the above verbs have an existential reading, they take one O argument, which is
either a noun phrase or a relative clause, as seen in (3.95-3.96), or take no O argument
when used in a monovalent copula construction, as seen in (3.97-3.98).
138
3.95
Tzimatsi aparoni yoimoshirinkiri.
tzimatsi aparoni y=oi-kimoshiri-Nk-i=ri
EXIST one
3m.A=CAUS-be.happy-NMZ-REAL=REL
‘There was one [idol] who made them rejoice [praise him].’
3.96
Tekatsi amanantatsini, tekatsi atziri.
tekatsi
amanaNt-atsi=ni
tekatsi
atziri
NEG.EXIST buy-PROG-STAT=REL NEG.EXIST person
‘There is no one who will buy [my fish], there are no people [buyers].’
3.97
Pairani ikanta konoya itsipatari ishamitzi.
pairani i=kaNt-a
konoya
before 3m.S=be-REAL tortoise
i=tsipat-a=ri
i=shamitzi
3m.S=accompany-REAL=3m.O
3m.poss=namesake
‘Long ago there was a tortoise with his namesake.’
3.98
-Tsika pikanta?-Kameetsataki.
tsika pi=kaNt-a
kameetsa-t-ak-i
WH 2S=be-REAL be.good-EP-PRF-REAL
‘How are you?’ ‘I am good.’
There is also a set of defective copula verbs such as the copula verb of existence na
‘be’, the negative copula of existence kaari, the copula of location saik ‘be at’, the copula
of naming pai, and the copula of capacity kara. A defining feature for a copula verb is
that it always occurs with two core arguments, copula subject (CS) and copula
complement (CC) (Dixon 2010a:160). By definition, copula verbs tend to be
morphologically irregular and show fewer forms than other verbs; they also tend to
belong to the semantic class of stative verbs such as stand, sit, exist, appear, become or
may derive from verbs of motion go or come (Payne 1997:117). The irregularity of
Ashéninka Perené copulative verbs is manifest in that they often lack subject proclitics
and don’t occur with most verbal categories. For example, the copula of existence na and
the copula of capacity kara are typically marked for subject, progressive aspect and realis
139
status whereas saik and pai frequently lack S-argument person markers; pai has shown
empty semantics and morphological irregularity when used in interrogative clauses. The
negative copula of existence kaari is an invariant form. A set of copulas in Ashéninka
Perené is given in Table 18.
TABLE 18. Copulas
Clause type
Equation
Attribution
Possession
Location
Other
Form
na, kaari
n, kaari
kaari
saik
pai, kaNt
kara
Function
express or deny identity
express or deny attribution
deny possession
express location
specify the name of an entity
express capacity
The copula of existence na is optional when it occurs with both nominal and
adjectival predicates, as shown in (3.99-3.101).
3.99
Noshinto ingeniero (onatzi).
no=shiNto
ingeniero o=na-tz-i
1SG.poss=daughter engineer 3n.m.S=be-EP-REAL
‘My daughter is an engineer.’
3.100 Noshinto kameetsanto (onatzi).
no=shiNto
kameetsa-Nto
1SG.poss=daughter be.good-ADJ.n.m.
‘My daughter is beautiful.’
3.101 Notomi matsari (inatzi).
no=tomi
matsa-ri
1SG.poss=son be.thin-ADJ.m.
‘My son is thin.’
o=na-tz-i
3n.m.S=be-EP-REAL
i=na-tz-i
3m.S=be-EP-REAL
The negative copula of existence kaari is used with nominal and adjectival predicates
and genitive constructions. Unlike the affirmative copula of existence na, which always
occurs clause-finally, kaari precedes the non-verbal predicate, as seen in (3.102-3.104).
140
3.102 Iriroite kaari ashaninka.
iriro-ite kaari
a=shaniNka
he-PL NEG.C 1PL.poss=fellowman
‘They are not our fellowmen.’
3.103 Kaari
matsari otsitzi.
NEG.C thin
dog
‘The dog is not thin.’
3.104 Kaarima ashi estado?
kaari=ma
ashi estado
NEG.C=DUB
its state
‘Isn’t it of the state [isn’t it a public school]?’
A distinctive feature of the negative copula constructions is that they are most often
found expressing an identity relation like in (3.102) and are rarely used in non-verbal
clauses expressing attribution and possession, exemplified by (3.103-3.104). Language
consultants point out that it is more common to use a negated verbal predicate to express
the relation of attribution, as seen in (3.105).
3.105 Te imatsate otsitzi.
te
i=matsa-t-e
otsitzi
NEG.REAL 3m.S=be.thin-EP-IRR dog
‘The dog is not thin.’
The language has two copulas of naming: pai and kaNt. They fill the slot of a
transitive verb, as seen in (3.106-3.107).
3.106 Tzimatsi aparoni atziri ipaita Naviriri.
tzimatsi aparoni atziri i=pai-t-a
Naviriri
EXIST one
person 3m.S=be.called-EP-REAL person’s.name
‘There was a person, he [who] was called Naviriri.’
141
3.107 Tzimatsi intakantarori isaikita aka, ikantziri Rasaro.
tzimatsi iNtakaNtarori i=saik-it-a
aka
EXIST ancestor
3m.S=be.at-before-REAL here
i=kaNt-tz-i=ri
Rasaro
3m.S=say-EP-REAL=3m.O person’s.name
‘Long ago there was an ancestor living here, they called him Lazar.’
The copula of naming pai shows morphologically defective behavior by exhibiting
few grammatical distinctions. It often lacks S personal proclitic and has minimal verbal
morphology. Examples are provided in (3.108-3.109).
3.108 -Tsika paita panteri?-Nooyakotatziro aparoni ariityanani novankoki.
tsika pai-t-a
p=aNt-e=ri
n=ooy-ako-t-atz-i=ro
WH be.called-EP-REAL 2S=do-IRR=REL 1SG.A=wait-EP-APPL-PROG-REAL=3n.m.O
aparoni arii-t-ia=na=ni
no=paNko-ki
one
arrive-EP-IRR=1SG.O=REL 1SG.poss=house-LOC
‘What are you going to do [what is it called that you are going to do]?’ ‘I am
expecting a female who will be visiting me in my house.’
3.109 -Yamakinari aparoni oisha. -Ninka? -Paitaranki…Antoko!’
y=am-ak-i=na=ri
aparoni oisha niNka
3m.A=bring-PRF-REAL=1SG.O=3m.O one
lamb who
pai-t-a=raNki
Antoko
be.called-EP-REAL=ADV.P person’s.name
‘He brought me a lamb.’‘Who?’ ‘Whachamacallit…Antoko!’
Similarly to other copulas, the copula of location saik does not inflect like regular
verbs. From dozens of available verb-specific categories, it tendentially selects the S
proclitic and suffixes of aspect and mode. Examples are given in (3.110-3.111).
3.110 -Tsika norirote?-Isaikatzi henokini mesaki.
tsika no=riro-te
i=saik-atz-i
henoki-ni
WH 1SG.poss=watch-poss 3m.S=be.at-PROG-REAL top-DIM
‘Where is my watch?’ ‘It’s on top of the table.’
mesa-ki
table-LOC
142
3.111 -Tsikama saikatsika?-Nonyakiro parenikinta.
tsika=ma saik-atsi=ka
no=ny-ak-i=ro
pareni-ki=Nta
WH=DUB be.at-STAT=Q 1SG.A=see-PRF-REAL=3n.m.O river-LOC=DEM
‘Where is she?’ ‘I saw her over there by the river.’
The copula of capacity kara also shows limited morphological possibilities in
comparison with regular verbs. It is typically found with the grammatical categories of
reality status (realis –i), and person (it is marked for subject), as seen in (3.112-3.113).
3.112 Te yayero okaratzi iyotakaivitaitarira.
te
y=ay-e=ro
o=kara-tz-i
NEG.REAL
3m.A=take.IRR-IRR=3n.m.O 3n.m.S=CAP.C-EP-REAL
i=yot-ak-a-vi-t-ai-t-a=ri=ra
3m.A=know-CAUS-EP-FRUS-EP-IMP.P-REAL=3m.O=ADV
‘He didn’t grasp all [material] which was taught to him.’
3.113 Tzimatsi otomi okaratzi mava.
tzimatsi o=tomi
o=kara-tz-i
mava
EXIST 3n.m.poss=son
3n.m.S=CAP.C-EP-REAL three
‘She had three children.’
3.4.3 Adjectives
Ashéninka Perené has two adjective sets, (i) a small, closed underived adjectives class
and (ii) a bigger, open derived adjectives class. The syntactic behavior and semantic
types of both adjective classes are similar whereas their morphological possibilities are
different.
(i) Underived adjectives. The tiny set of underived adjectives covers the meanings of
size, age, quantification, and similarity. The list of these adjectives is given in Table 19.
Following Dixon’s typology of adjectives, Ashéninka Perené underived adjectives can be
classified as a mixed type which combines some of the properties of nouns with some of
143
the verbs (2010a:63). Adjectives occur in NPs, inflecting like a noun, and also as heads of
transitive and intransitive predicate, inflecting like a verb.
The basic position for a modifying adjective in an NP is before the noun head. When
underived adjectives have grammatical properties similar to those of nouns, they take
nominal inflectional markers such as gender (masculine –ri, non-masculine –ro), case
(locative –ki), and number (plural number –paye). Examples are provided in (3.1143.116).
TABLE 19. A set of underived adjectives
Semantic type
Size
Age
Quantification
Similarity
Form
aNtari/ro
iyaani
antari/ro
iyaani
evaNkari/ro
kiNkivari/ro
maaroni
osheki
kapichee
aparoni
apite
mava
pashini
Gloss
‘big’
‘small (of size)’
‘mature (of age)’
‘young (of age)’
‘adolescent’
‘old’
‘all’
‘many’
‘few’
‘one’
‘two’
‘three’
‘other’
3.114 Iyovaritakotakina apa antaro korarintsi.
i=kovari-t-ako-t-ak-i=na
apa aNtaro korarintsi
3m.A=corral-EP-APPL-EP-PRF-REAL=1SG.O father big.n.m corral.n.m.
‘My father built a big corral for me.’
3.115 Nantavaitaki osipitariki antaropaye.
n=aNta-vai-t-ak-i
osipitari-ki
aNtaro-paye
1SG.S=work-DUR-EP-PRF-REAL hospital-LOC big.n.m-PL
‘I worked in big hospitals.’
3.116 Isaiki antaroki inchato.
i=saik-i
aNtaro-ki
inchato
3m.S=be.at-REAL big.n.m-LOC tree
‘It was in the big tree.’
144
As far as nominal derivational morphology is concerned, underived adjectives can
take classifiers and degree (augmentative and diminutive) suffixes. Examples are
provided in (3.117-3.120).
3.117 Pincheke antaroshi kompiroshi.
pi=N-chek-e
aNtaro-shi
koNpiro-shi
2S=IRR-cut-IRR big-CL:small.thin.flexible
palm.sp.-leaf
‘Cut the big leaves of the palm tree (Phytelephas microcarpa).’
3.118 Yantzi antarote pankotsi.
y=aNt-tz-i
aNtaro-ite paNkotsi
3m.S=make-EP-REAL big-AUG house
‘They made an immense house.’
3.119
Antarosante piyempita!
aNtaro-saNte pi=yeNpita
big-AUG
2poss=ear
‘You have a big ear [you don’t listen]!’
3.120 Tzimatsi neentsite, tzimatsi iyaanini shinkantzi.
tzimatsi n=eentsi-te
tzimatsi iyaani-ni
shiNka-ntzi
EXIST 1SG.poss=child-poss EXIST small-DIM cry.a.lot -ADJ.m.
‘I have a baby, a tiny cry-baby.’
Underived adjectives can function as arguments, without a nominal head. The
nominal head can be either inferred from the context or may be mentioned in previous
discourse. An example of an adjective used without a noun head is given in (3.121).
3.121 Opinatziro kapicheeni.
o=pina-tz-i=ro
kapichee-ni
3n.m.A=pay-EP-REAL=3n.m.O small-DIM
‘She pays them little [money].’
145
Underived adjectives also have properties similar to those of verbs. They can function
as heads of intransitive and transitive predicates, taking marking of person, aspect, and
realis status, as seen in (3.122-3.123).
3.122 Aparoni kooya ovakera antarotatzi.
aparoni kooya ovakera aNtaro-t-atz-i
one
woman recently 3n.m.S.big-EP-PROG-REAL
‘A woman who menstruated for the first time.’
3.123
Airo pishikitziro.
airo
pi=osheki-tz-i=ro
NEG.IRR 2A=many-EP-REAL=3n.m.O
‘Don’t pick many [oranges].’
The criteria for recognition Ashéninka Perené adjectives as a separate class include
two functions which they perform: (i) as a modifier within the NP and (ii) as a predicate
adjective (attributive). Examples of these functions are seen in (3.124-3.125).
3.124 Nonyatziro aparoni ochevo, pashini.
no=ny-atz-i=ro
aparoni ochevo pashini
1SG.A=see-PROG-REAL one
branch another
‘I saw one branch, another [branch].’
3.125 Inkaari antaro.
lake big
‘The lake is big.’
The property that makes underived adjectives distinct is that they take some
adjective-specific derivational morphology such as the augmentative suffix –ni, as shown
in (3.126).
146
3.126 Yaminapakiri oshekini ashaninka.
y=amin-ap-ak-i=ri
osheki-ni
3m.S=see-DIR-PRF-REAL=3m.O many-AUG
‘A lot of our fellowmen saw him.’
a=shaniNka
1PL.poss=fellowman
The properties of underived adjectives are summarized in Table 20.
TABLE 20. Properties of underived adjectives
Morphology
Syntax
Semantics
Noun-like properties
Verb-like properties
gender markers –ri/ro
plural marker –paye
locative case marker -ki
diminutive –(i)ni,
augmentative –ite, -saNte
classifier -shi
can be used
independently without a
nominal head
person markers
aspect markers
mode markers
function as heads of
intransitive and transitive
predicates
Unique or adjectivelike properties
augmentative -ni
function as modifiers of
nominal heads
and as predicate
adjectives
state that something has
a certain property or
specify the noun
referent; cover the most
prototypical semantic
fields associated with
adjectives such as
dimension, age, color,
etc.
(ii) Derived adjectives. An open subclass of adjectives, derived from verb roots,
nominalized verbs, nouns, and adverbs, takes a set of gender-specific adjectivizing
suffixes such as -ntzi/-Nto, -tsiNkari/-tsiNkaro, and –ri/ro. Another common adjectivizing
technique is to combine the relativizing/nominalizing/adjectivizing form –ri, which is not
gender-sensitive, with a verb. The class of derived adjectives is associated with the
following semantic categories (based on the categorization guidelines proposed by Dixon
2010a:73-4).
147
1. AGE: ovakerari ‘new, recent’
2. VALUE: kameetsari/ro ‘good’, kameetsantzi/Nto, kaNtaperantzi/Nto ‘beautiful’,
katsimari ‘bad’
3. COLOR: kityoNkari/ro ‘red’, kitamarori/kitamaro ‘white’, natsiryari/ro ‘green’,
potsitari/ro, kisaari ‘black’, cheeNkari ‘obscure’, kiteriri ‘yellow’, samaNpo ‘grey’,
‘ashen’
4. PHYSICAL PROPERTY: matsari/ro ‘thin’, santsari ‘long’, shintsiri ‘strong, robust’,
taNpatsikari ‘straight’, tsaiNpiri ‘sharp’, tenari ‘heavy’, katsiNkari ‘cold’, saavaari
‘hot’, kipishiri ‘bitter’, pochari ‘sweet’, kachori ‘acid’, katyori ‘salty’, masaNkari
‘bland’, anatzirori ‘raw’, posari ‘cooked’, mantsiyari/ro ‘sick’, kamaiveNtsiNkari/ro
‘widowed’
5. HUMAN PROPENSITY: yotanitsiNkari/ro ‘ smart’, aNtavairintzi/Nto ‘hard-working’,
shiroNtaryantzi/nto ‘cheerful’
6. SPEED: shintsiri ‘quick’
7. QUALIFICATION: iripirori/ iropirori ‘correct’
The adjectives from this subclass have grammatical properties similar to those of
nouns and nominalized verbs. Their main function is modification of NPs. As far as NP
constituent order is concerned, the modifier’s position is typically after the head noun. An
important feature of such modified nominal phrases is that the head nouns are frequently
omitted, with the derived adjectives functioning as arguments in such cases. Examples of
a modified headed NP and an anaphoric NP with an adjective as a sole lexeme are
provided in (3.127-3.128).
148
3.127 Nontyoonkanakyari ashininka mantsiyari.
no=N-tyooNk-aNt-ak-ia=ri
a=sheniNka
1SG.A=IRR-bathe-APPL.REAS-PRF-IRR=3m.O 1PL.poss=fellowman
mantsiya-ri
be.sick-ADJ.m
‘I’ll bathe in steam this sick fellowman.’
3.128 Novayetziri inchashi nonkaatantyariri mantsiyari.
no=v-a-ye-tz-i=ri
inchashi
1SG.A=put-EP-DIST-EP-REAL=REL herb
no=N-kaa-t-aNt-ia=ri=ri
mantsiya-ri
1SG.A=IRR-bathe-EP-APPL.REAS-IRR=3m.O=REL be.sick-ADJ.m
‘The leaves that I put here are for treating the sick [person].’
Derived adjectives may function as predicate adjectives in verbless clauses, as seen in
(3.129-3.130). When adjectives occur in copula clauses, as in (3.131), they are positioned
clause-initially. Language consultants note that modified NPs in copula clauses indicate
emphasis in the statement of property.
3.129 Tsipana natsiyaro.
tsipana natsiya-ro
leaf
be.green-ADJ.n.m
‘The leaves are green.’
3.130 Iroo kameetsanto, irotaki yotanitsinkaro, notomi iritaki yotanitsinkari.
iroo kameetsa-Nto
irotaki yotani-tsiNkaro no=tomi
she be.good-ADJ.n.m FOC know-ADJ.n.m. 1SG.poss=son
iritaki yotani-tsiNkari
FOC know-ADJ.m
‘She is pretty, she is clever, my son is clever [too].’
3.131 Kityonkaro onatzi kitsarintsika.
kityoNka-ro
o=n-atz-i
kitsa-rintsi=ka
be.red-ADJ.n.m 3n.m.S=be-PROG-REAL dress-NOM=DEM
‘The cushma [man’s traditional robe] is red.’
149
Similarly to nouns, the derived adjectives take nominal inflectional (plural marker
–paye) and derivational morphology (classifying derivational suffix –ni ‘deceased
referent’, diminutive –(patsa)ini, augmentative –ni). Examples are seen in (3.132-3.135).
3.132 Okanta aminantariri mantsiyaripaye.
o=kaNt-a
amin-aNt-a=ri=ri
3n.m.S=happen-REAL treat-APPL.REAS-REAL=3m.O=REL
mantsiya-ri-paye
be.sick-ADJ.m-PL
‘So that we [could] treat the sick.’
3.133
Nokimavahiro inani, anini kamahatsirini.
no=kim-av-ah-i=ro
ina-ni
ani-ni
1SG.A=listen-DIR-REGR-REAL=3n.m.O mother-deceased grandpa-deceased
kam-ah-atsi-ri-ni
die-REGR-STAT-ADJ-deceased
‘I listened to my late mother and grandfather.’
3.134 Tenarini notsima.
tenari-ni
no=tsima
be.heavy.ADJ-AUG 1SG.poss=wood
‘My wood is too heavy.’
3.135 Irotaki nosaatantari mantsiyaripatsaini.
irotaki no=saa-t-aNt-a=ri
FOC
mantsiya-ri-patsa-ini
1SG.A=steam.bathe-EP-INST-REAL=3m.O be.sick-ADJ.m-small.part-DIM
‘With this I bathe the sickly [little things].’
Properties of derived adjectives are summarized in Table 21.
Comparative strategies. Ashéninka Perené has a number of non-canonical comparative
strategies to indicate a contrast or similarity between two entities. When two entities are
contrasted, the standard NP is not expressed; only the comparee NP occurs24.
24
Stassen states that a comparative construction usually involves three elements: a gradable predicate and
two objects, typically encoded as two noun phrases (comparee NP and standard NP) (2008).
150
TABLE 21. Properties of derived adjectives
Morphology
Syntax
Noun-like properties
gender-specific markers –ri/ro,
-ntzi/-nto, -tsinkari/tsinkaro
diminutive –(patsa)ini
classifying suffix –ni ‘deceased’
plural -paye
can be used independently without a
nominal head
Semantics
Unique or adjective-like properties
augmentative -ni
adjectivizer -ri
function as modifiers of nominal
heads, complements in copula clauses,
and predicate adjectives
state that something has a certain
property or specify the noun referent;
cover the most prototypical semantic
fields associated with adjectives such
as age, color, physical property, etc.
The gradable scale is marked on predicate by the verbal derivational intensifier/degree
marker suffix –pero~-piro, as seen in (3.136-3.137).
3.136 Iroka maaakinaka antaroperotatzi.
iroka
maakina=ka
aNtaro-pero-t-atz-i
DEM.n.m machine=DEM
big-DEGR-EP-PROG-REAL
‘This machine is bigger (or the biggest).’
3.137 Iroka inchatoka iro antaroperotatzi.
iroka
inchato=ka
iro aNtaro-pero-t-atz-i
DEM.n.m tree=DEM
but big-DEGR-EP-PROG-REAL
‘This tree is the biggest of all.’
Another common comparative strategy is to use the verb shiy ‘to be like’ as the
gradable predicate in positive or negative polarity clauses. In (3.138), the comparee is
expressed as the person proclitic on the gradable predicate whereas the standard NP
occurs in reduced form. The gradable verb predicate shiy ‘be like’ is marked for the
comparee NP (by 3n.m. person clitic o=) and the standard NP (by the O-oriented goal
directional which codes semantic patients), as seen in (3.139). In a conjoined comparative
construction, shown in (3.140), which consists of two structurally independent clauses,
151
the first clause contains the comparee NP tyaapaka ‘this hen’ and the second -the
standard NP varipa ‘rooster’.
3.138 Te oshiyaro pashinira.
te
o=shiy-ia=ro
pashini=ra
NEG. REAL 3n.m.A=be.like-IRR=3n.m.O other=DEM
‘This one isn’t like that one [they aren’t the same].’
3.139 Iroka maakinaka ari oshiyavakahetani/oshiyavakaheta/oshiyavaka.
iroka
maakina=ka
ari o=shiy-av-ak-a-he-t-a-ni
DEM.n.m machine=DEM PP 3n.m.A=be.like-DIR-PRF-EP-PL-EP-REAL-PL
‘This machine is like that one [they are the same].’
3.140 Tyaapaka te ontenapirote, te oshiyari varipa.
tyaapa=ka te
o=N-tena-piro-t-e
te
hen=DEM NEG.REAL 3n.m.S=be.heavy-DEGR-EP-IRR NEG.REAL
o=shiy-ia=ri
varipa
3n.m.A=be.like-IRR=3m.O rooster
‘The hen doesn’t weigh as much as the rooster, they are not the same.’
Additionally, a set of personal pronouns can be used to convey contrast (see §3.4.5
for details). An example of the personal pronoun irintsi ‘but he’, which has contrastive
‘exhaustive listing’ only-emphasis, is provided in (3.141).
3.141 Koki irintsikya tzimakotatzi iyaniri.
koki
irintsi=kya
tzim-ako-t-atsi
i=kaniri
father-in-law he.FOC=EMPH exist-APPL-EP-STAT 3m.poss=manioc
‘My father-in-law is the only one who grows his manioc [literally ‘exists with
manioc’].’
3.4.4 Adverbs
By definition, adverbs are a catch-all category which covers a wide range of concepts
(Payne 1998:69). Based on their semantic content, Ashéninka Perené adverbs can be
152
grouped into a few classes such as (i) time adverbs, (ii) place adverbs, (iii) locativeexistential adverbs, and (iv) miscellaneous others. Adverbs are too heterogeneous to be
defined by category-specific morphology, although they often take word-final suffixes –
ni and –ra, e.g. pairani ‘before’, okaakini ‘in close proximity’, tsiteni ‘at night’, ovakera
‘recently’. Most adverbs exhibit obscure internal morphology, which makes identification
of their sources problematic. In terms of their distribution, adverbs can appear freely in
any position in the clause. I will discuss each adverb subclass in detail below.
(i) Time adverbs. As shown in Table 22, a number of verbs, e.g. pony ‘come from’, kaNt
‘happen’ function as temporal adverbs. Some temporal adverbs are derived from verbs,
e.g. sheete ‘be.afternoon ’> sheeteni ‘in the afternoon’, tsite ‘be dark’ > tsiteni ‘at night’;
or from an adjective-verb compound kapichee-kitaite [small-be.morning]>
kapichekitaiteni ‘early in the morning’; from a numeral aparoni ‘one’> aparohatzini ‘one
day’.
There are also noun phrases and whole clauses which function as adverbial
expressions. Such expressions are used in an adverbial role, as modifiers of the verb
phrase. Examples of adverbial expressions are provided in (3.142-3.143).
3.142 Kamaki aparoni kooya pasinikiranki osarentsi.
kam-ak-i
aparoni kooya pasini-ki=raNki
osarentsi
die-PRF-REAL one
woman other-LOC=ADV.P year
‘One woman died last year.’
3.143 Mavatapaintziri inyatziri amemporipaye.
mava-t-apaiNt-tz-i=ri
i=ny-atz-i=ri
ameNpori-paye
three-EP-once-EP-REAL=REL 3m.A=see-PROG-REAL=3m.O vulture-PL
‘Three days later, they saw vultures.’
153
TABLE 22. Temporal adverbs
Form
ovakera(ni)
osaikera
osaitekera
ironyaaka
ironya
irohatzi
pairani
apaata
ponyashita(ka)
Gloss
recently
tomorrow
tomorrow
now
now
until
before, since
later, soon
afterwards
iNponya
i/oponya
ponyaka
iponyashitanaka
afterwards
(IRR)
afterwards
afterwards
afterwards
aparohatzini
osamani
koramani
kapichekitaite(ni)
tsiteniraNki
sheteniraNki
sheteni
tsiteni
chapiNki(raNki)
iNkaraNki
i/okaNta
once, one day
a while (ago)
in the future
early morning
last night
yesterday
in the afternoon
at night
the other day
some time ago
meanwhile
Origin
o=v-ak-e=ra [3n.m.S=put?-PRF-REAL=ADV]
o=saik-e=ra [3n.m.S=be.at-REAL=ADV]
o=sa<ite>k-e=ra [3n.m.S=be.at<AUG?>IRR=ADV]
iro-hatzi [precisely-also]
pai-ra-ni [?-ADV-ADV]
pony-ashi-t-ak-a [come.from-APPL.INT-EP-PRFREAL]
i=N-pony-a [3m.S=IRR-come.from-IRR]
i/o=pony-a [3m./n.m.S=come.from-REAL]
pony-ak-a [come.from-PRF-REAL]
i=pony-ashi-t-an-ak-a [3m.S=come.from-APPL.INTEP-DIR-PRF-REAL]
aparo<hatzi>ni [one<also>ADV]
kapichee-kitaite(-ni) [small-morning-ADV]
tsiteni=raNki [night=ADV.P]
sheteni=raNki [day=ADV.P]
i/o=kaNt-a [3m./n.m.S=happen-REAL]
3.144 Osamanityanaki ihatashitakiri iraniri ivankoshiki.
osamani-tya-an-ak-i
i=ha-t-ashi-t-ak-i=ri
late-almost-DIR-PRF-REAL 3m.A=go-EP-APPL.INT-EP-PRF-REAL=3m.O
ir=ani-ri
i=paNko-shi-ki
3m.poss=brother-in-law-poss 3m.poss=house-CL:small.thin.flexible-LOC
‘Some time later, he approached his brother’s-in-law hunting shack.’
Temporal adverbs are clausal adjuncts and typically occur clause-initially or clausefinally. They can also precede or follow the predicate, as seen in (3.145-3.146).
154
3.145 Ikantapaintziri chapinki: -300 soles.
i=kaNt-apaiNt-tz-i=ri
chapiNki
300 soles
3m.A=say-once-EP-REAL=3m.O the.other.day 300 Peruvian.currency
‘He said to him the other day, ‘[I’ll pay you] 300 soles.’
3.146 Ishimatepaye inkaranki akiparyakotaki.
i=shima-paye
iNkaraNki
akipa-ry-ako-t-ak-i
3m.poss=fish-PL some.time.ago wrap-REV-APPL-EP-PRF-REAL
‘They removed the leaves off the cooked fish a while ago.’
Bound temporal adverbial forms. The following three formatives derive adverbs from
nouns and verbs: –paite ‘during’, e.g. iNkani-paite [rain-during] ‘during the rainy
season’; -ki ‘when, during’ (note that the spatial reference of this locative suffix is
extended to time), e.g. kitaiteri-ki [day-LOC.T] ‘during the day’; and =raNki ‘in the
past’, e.g. savadoranki savado=ranki [Saturday=ADV.P] ‘last Saturday’. Examples are
provided in (3.147-3.149).
3.147 Oshiki notashi tampatsikapaite.
oshiki no=tash-i
taNpatsika-paite
many 1SG.S=be.hungry-REAL be.noon-ADV
‘I am very hungry at noon.’
3.148
Ahate katsinkaiteriki.
a=hat-e
katsiNkaite-ri-ki
1PL.S=go-IRR be.cold-ADJ-LOC
‘We’ll go when it’s cold.’
3.149 Nonyavakiro hinokitapakiranki oryaatsiri pava.
no=ny-av-ak-i=ro
hinoki-t-ap-ak-i=raNki
1SG.A=see-DIR-PRF-REAL=3n.m.O up-EP-DIR-PRF-REAL=ADV.P
ooryaatsiri pava
sun
solar.deity
‘I saw it when the sun was high.’
155
(ii) Place adverbs. Direction/location or place adverbs, like temporal adverbs, cannot take
any noun morphology, or be possessed or modified. They cannot function as arguments.
Place adverbs are often formed with the help of the demonstrative local adverbial clitics
=ka ‘proximal’, =ra ‘medial’, =Nta ‘distal’. The set of place adverbs includes
demonstrative adverbs aka~haka ‘here’, aNta~haNta ‘there’, yoNta ‘over there.’ There is
also a subset of spatial adverbs okaakini ‘in close proximity’, iNtaina ‘far away’,
intsoNpaikira/intsoNpoikiNta ‘inside’, iNtakiroNta ‘outside’, iNtatzikiroka ‘this side of
the river’, iNtatzikiroNta ‘the other side of the river’, katoNko ‘upriver’, kiriNka
‘downriver/central part of the jungle’. Examples are given in (3.150-3.153).
3.150 Ikenashihetanaha pashini kirinka, pashini katonko.
i=ken-ashi-he-t-an-ah-a
pashini kiriNka
pashini katoNko
downriver other upriver
3m.S=go-APPL.INT-PL-EP-DIR-REGR-REAL other
‘Some went downriver, others went upriver.’
3.151 Ikinakanakina intsompoikinta,ikinakanakina intakironta.
i=kin-ak-an-ak-i=na
intsoNpoi-ki=Nta
3m.A=walk-CAUS-DIR-PRF-REAL=1SG.O inside-LOC=ADV
i=kin-ak-an-ak-i=na
3m.A=walk-CAUS-DIR-PRF-REAL=1SG.O
‘He walked with me inside, then outside.’
iNtakiro=Nta
outside=ADV
3.152 Ari osaikapakeri misión intatzikironta.
ari o=saik-ap-ak-i=ri
misión iNtatzikiroNta
PP 3n.m.S=be.at-DIR-PRF-REAL=REL mission this.side.of.the.river
‘There was a[n adventist] mission founded on this side of the river.’
3.153 Chooki osaiki okaakini.
chooki o=saik-i
okaakini
sister 3n.m.S=be.at closeby
‘My sister lives in close proximity.’
156
(iii) Locative-existential adverbs.A closed subclass of gender-sensitive locativeexistential adverbs is organized in two gender-based paradigmatic sets expressing a threeway contrast in terms of distance from the speaker. Locative-existential adverbs in Table
23 are formed by suffixing the proximal –ka, medial –ra and distal –Nto to the
pronominal stems hiri- and hiro- derived from the third person singular personal
pronouns iriro (3SG.m) and iroo (3SGn.m).
TABLE 23. Locative-existential adverbs
Gender
Masculine
Non-masculine
Proximal
hirika
hiroka
Medial
hirira
hirora
Distal
hiriNto
hiroNto
Sometimes bare pronominal stems are used in natural discourse. Their function is
two-fold: to make spatial reference to an entity and express an invitation to the
interlocutor to make use of it. The gender assignment of the locative-existential adverbs
is based on the gender class of the proffered noun referent. In (3.154, 3.156), the animate
referent of noimiporokia ‘that husband of mine’ and maniro ‘deer’ is of masculine
gender; hence, the use of hiri ‘here he/it [3pSGm] is’. In (3.155), the referent of the
nominal stem paNko ‘house’ is inanimate, by default taking agreement as a nonmasculine entity. In accordance with agreement rules, hiroka ‘here it [3pSGn.m.] is’ is
selected.
3.154 Noimiporokira hiri isaikaporokitzira.
no=imi-poroki=ra
hiri
1SG.poss=husband-CL:quantity=DEM there.he.is
i=saik-a-poroki-tz-i=ra
3m.S=be.at-EP-CL:quantity-EP-REAL=ADV
‘My husband, there he is, he sits there, a bag of bones [you can talk to him].’
157
3.155 Hiroka pivankoka.
hiroka
pi=paNko=ka
here.it.is 2poss=house=DEM
‘Here it is, your house [you can use e.g. sleep there].’
3.156
Noina, hirika ivatsa maniro, pinkotsitavake.
no=ina
hirika
i=vatsa
maniro pi=N-kotsi-t-av-ak-e
2S=IRR-cook-EP-DIR-PRF-IRR
1SG.poss=wife here.it.is 3m.poss=meat deer
‘Wife, here it is, deer meat, cook it.’
(iv) Other adverbs. This small heterogeneous group includes aspect, modal, manner, and
degree adverbs, among others. The position of these adverbs largely depends on their
function. When the adverb modifies the predicate, it either precedes or follows the
predicate. Aspect adverbs aikiro ‘still’ and tekira ‘not yet’ typically precede the verb,
while other adverbs such as apaniroini ‘alone, solely’, shintsini ‘loud’, ‘fast’, neetsini
‘slowly’, and degree (scalar) adverbs osheki ‘a lot’ and kapicheeni ‘little’ can occur either
before or after the predicate, as shown in (3.157-3.160).
3.157 Tekira omposate piyaniri.
tekira o=N-posa-t-e
not.yet 3n.m.S=IRR-cook-EP-IRR
‘Your manioc has not cooked yet.’
pi=kaniri
2poss=manioc
3.158 Aikiro oshitovatzi nihaa.
aikiro o=shitov-atz-i
nihaa
still 3n.m.S=leave-PROG-REAL water
‘The water level keeps going down.’
3.159 Osheki nonintzi nonyaatsatya.
osheki no=niNt-tz-i
no=nyaatsa-t-ia
a.lot 1SG.S=want-EP-REAL 1SG.S=play-EP-IRR
‘I liked to play a lot.’
3.160 Nareetantahari shintsini.
n=aree-t-aNt-ah-a=ri
shintsi-ni
1SG.S=arrive-EP-APPL.REAS-REGR-REAL=REL be.strong-ADV
‘That’s why I came back quickly.’
158
When the adverb functions on the clausal level, it is positioned at the clause
periphery. Modal adverbs occur clause-initially. They have complex internal morphology
and are derived from the positive polarity complement-taking verb ari ‘be the case’. In
single-clause statements, ari denotes ‘all right, OK’. The ari-derived adverbial forms
express the following modal senses: arima ari=ma [PP=DUB] ‘perhaps’, aritaima
ari=taima [PP=DUB] ‘perhaps, maybe’, arikyaro ari=kya=ro [PP=EMPH=3n.m.O]
‘that’s true, indeed’. Examples are provided in (3.161-3.163).
3.161 Arikyaro, nominkatakaro, notovantarori.
ari=kya=ro
no=miNka-t-ak-a=ro
PP=EMPH=ADV 1SG.A=make.a.hunting.blind-EP-PRF-REAL=3n.m.O
no=tov-aNt-a=ro=ri
1SG.A=fell-APPL.REAS-REAL=3n.m.O=REL
‘That’s right, I made the platform to hunt animals, that’s why I had felled it [the
tree].’
3.162 Aritaima, osaikera kapichikitaite nataite.
ari=taima osaikera kapichi-kitaite
n=atai-t-e
PP=DUB tomorrow small-be.morning 1SG.S=go.up-EP-IRR
‘Perhaps, tomorrow early in the morning, I’ll go up [the hill to do the job].’
3.163
Arima irakero iraantsi?
ari=ma
ir-ak-e=ro
iraa-ntsi
PP=DUB drink-PRF-IRR=3n.m.O blood-ABS
‘Perhaps she’ll drink blood.’
Adverbs often take emphatic marking such as exclamative (general focus) =ve,
focused object (?) =kya, and contradictory emphasis=ha. Examples are provided in
(3.164-3.166).
3.164 Ironyakya ivaitakiri.
ironya=kya i=v-ai-t-ak-i=ri
now=EMPH 3m.A=kill-IMP.P-EP-PRF-REAL=3m.O
‘Finally, he was killed.’
159
3.165 Kapicheeni ironyakaha kamatsavaitahina intakiro.
kapicheeni ironyaka=ha kamatsa-vai-t-ah-i=na
iNtakiro
little
now=EMPH crawl-DUR-EP-REGR-REAL=1SG.O outside
‘Little by little I began to crawl to the area outside.’
3.166 Yookaitakina intainave!
y=ook-ai-t-ak-i=na
iNtaina=ve
3m.A=throw.away-IMP.P-EP-PRF-REAL=1SG.O far.away=EXCL
‘They threw me far away.’
3.4.5 Pronouns
This section deals with the closed subclasses of pronouns such as (i) personal, (ii)
possessive, (iii) demonstrative, (iv) interrogative, and (v) indefinite. (Ashéninka Perené
has neither reflexive nor relative pronouns.) I’ll discuss each subset of pronouns below.
(i) Personal pronouns. They are defined as ‘shifters whose reference relates to
participants in the speech act’ constituting ‘a small closed class of grammatical words
which vary for person’ (Dixon 2010a:189). Among four sets of personal pronouns,
attested in the language, one of them is identified as containing morphologically simple,
non-obligatory forms which function as discourse topics. The simple personal pronouns
are distinguished for person, number, and gender in 3p singular, as shown in Table 24.
Plural forms of pronouns are formed with the augmentative marker –ite. Non-singular
person may be of two types, inclusive (including the addressee) and exclusive (not
including the addressee). The inclusive/exclusive distinction is maintained only by older
speakers whereas younger speakers do not use the exclusive form naakaite, considering it
archaic. Examples are provided in (3.167-3.168).
160
TABLE 24. Forms of simple personal pronouns
Person/Gender
Number
singular
1
2
3 masculine
3 non-masculine
naaka
aviroka
iri(ro)
iroo
plural
exclusive
naakaite
avirokaite
iriroite~iririroite
iroite
inclusive
aroka~arokaite~arorite
3.167 Naakaite nosaikahetzi aka.
naakaite
no=saik-a-he-tz-i
aka
1PL.EXCL 1SG.S=be.at-EP-PL-EP-REAL here
‘We, we live here.’
3.168 Irooite amahetatzi nihaaki.
irooite
ama-he-t-atz-i
nihaa-ki
3PL.n.m. swim-PL-EP-PROG-REAL river-LOC
‘The women, they are swimming in the river.’
In addition to the pronominal set from Table 24, there are three more sets of personal
pronouns, each performing different pragmatic functions. A summary of the four sets is
given in Table 25. The category of number is not included in Table 25 since only
continuous discourse topic pronouns (set 1) are found in the plural number, formed with
the marker of plurality –ite (see Table 24 above). As shown in Table 25, Ashéninka
Perené has paradigmatic sets of topic and focus pronouns. Types of focused pronouns are
distinguished on the basis of the scope of focus (i.e. the scope of focus is limited to a
single constituent, be it subject or another verb argument). Types of topic pronouns are
differentiated on the basis of contrast (i.e. contrastive topic refers to one topical referent
explicitly opposed to another).
161
Table 28. Four sets of personal pronouns
Person/Gender/Number Set 1
Continuous
discourse
topic
1 singular
2 singular
3masculine singular
3non-masculine
singular
1 plural
naaka
aviroka
iri(ro)
iroo
Set 2
Contrastive
additive listing
focus on
S/A,O
naari
avirori
irirori
iroori
Set 3
Contrastive
exhaustive
listing focus on
S/A, O
naakataki
avirotaki
iritaki
irotaki
aroka
arori
arotaki
Set 4
Contrastive
discourse topic
naintsi
avintsi
irintsi
irontsi
Set 1. The first set of personal pronouns in Table 25 is used to signal a continuous
discourse topic. Topic pronouns are optional and are co-referential with the subject
proclitics. Note the distinctive position of the pronouns, found immediately to the left of
the verb in (3.169-3.170).
3.169 Iriro ikotsivintziro itsipatari Pito ironyaaka.
iriro i=kotsi-viNt-tz-i=ro
i=tsipat-a=ri
Pito ironyaaka
he 3m.A=cook-BEN-EP-REAL=3n.m.O 3m.A=join-REAL=3m.O name now
‘He, he cooks for them with Pito.’
3.170 Iroo pikante osankinatemita.
iroo pi=kaNt-e o=saNkina-t-e=mi=ta
she 2A=say-IRR 3n.m.A=write-EP-IRR=2O=OPT
‘She, she will write down what you say.’
Set 2. The personal pronouns from set 2 typically appear to the right of the verb signaling
a focused participant; they also occur in the preverbal position. The pronouns are derived
from the topic pronouns by the suffixation of the formative –ri. The focused pronouns
express contrastive ‘expanding’ emphasis on a subject or object constituent when the
speaker believes that the addressee has no knowledge of this particular participant’s
additional role in the described event; e.g. Doctors prescribe medicines, I also cure with
162
them. Examples of this type of contrast, which is called here ‘additive listing’, are
provided in (3.171-3.173).
3.171 Tekatsi naye naari pashini.
tekatsi n=ay-e
naari pashini
EXIST 1SG.S=take-IRR I.FOC another
‘There is nothing else which I will also use.’
3.172 Ikantzi irirori, ‘Te noyotero.’
i=kaNt-tz-i
irirori te
no=yo-t-e=ro
3m.S=say-EP-REAL he.FOC NEG.REAL 1SG.A=know-EP-IRR=3n.m.O
‘He also said, ‘I don’t know [how to make it].’
3.173 Amaki iroori, tzimatsi ivetsikakotakiro.
am-ak-i
iroori
tzimatsi i=vetsik-ako-t-ak-i=ro
bring-PRF-REAL she.FOC EXIST 3m.A=make-APPL-EP-PRF-REAL=3n.m.O
‘She also brought, she has a document [made by the university].’
Set 3. The third pronominal set consists of verbalized pronouns naakataki, avirotaki,
iritaki, and irotaki. The verbalized pronouns are formed from the topic pronouns naaka,
aviroka, iri(ro) and iroo and are inflected for aspect and realis status: -t-ak-i [EP-PRFREAL].This set is used to signal a focused subject, object of a transitive clause, or a
focused oblique constituent promoted via applicative derivation to the status of direct
object. This focus type expresses ‘exhaustive listing’ exemplified by the following
English sentence: He ate only the pulp of pifayo. The position of the focused verbalized
pronouns is at the left periphery of the clause. The selection of the gender-specific
3person singular verbalized pronouns iritaki/irotaki occurs in conjunction with the gender
class of the focused verb argument. When the referent of the third person noun referent is
masculine, iritaki is used; in other cases, the focus on the referent NP is expressed by
irotaki; naakataki and avirotaki are used when the referent of the focused NP in S/A
163
function is first person singular and second person respectively. Multiple marking of the
focused NP is possible: the focused verbalized pronouns often appear with the adverbs
apaniroini ‘on one’s own, solo’, intaani ‘only’, or the numeral aparoni ‘one’ to
emphasize the exhaustive listing effect. Note that co-referential verbal person clitics are
elided when human subjects are focused; verbal person clitics in O function do not delete
when objects are focused; obliques (e.g. those coded by the applicative of reason -aNt)
are typically not marked on the verb. Examples of focused peonouns are seen in (3.1743.184).
TRANSITIVE SUBJECT
3.174 Intaani nosaryanikipatsaini, iritaki naari kimoshiritakanari.
intaani no=sari-aniki-patsa-ini
iritaki naari
only 1SG.poss=grandma-DIM-small.part-DIM FOC I.FOC
kimoshiri-t-ak-a=na=ri
be.joyous-EP-CAUS-REAL=1SG.O=REL
‘ONLY my grandchildren, THEY ARE THE ONES WHO make me happy.’
3.175 Naakataki aakiro.
naakataki a-ak-i=ro
FOC
take-PRF-REAL=3n.m.O
‘I took it.’
INTRANSITIVE SUBJECT
3.176 Iritaki saikatsi aparoni.
iritaki saik-atsi
aparoni
FOC be.at-STAT one
‘HE IS THE ONE who is living here.
3.177 Irotaki ontsitenitanake, maperotanakina.
irotaki o=N-tsiteni-t-an-ak-e
mapero-t-an-ak-i=na
FOC 1SG.S=IRR-get.dark-EP-DIR-PRF-IRR get.worse-EP-DIR-REAL=1SG.O
‘IT is about to get dark, I am getting worse.’
164
OBJECT OF TRANSITIVE CLAUSE
3.178 Irotaki yaapakiro apa, osaikaperotantapakari Marankiaroki.
irotaki y=a-ap-ak-i=ro
apa
FOC 3m.A=take-DIR-PRF-REAL=3n.m.O father
o=saik-a-pero-t-aNt-ap-ak-a=ri
Marankiaro-ki
3n.m.S=be.at-EP-AUG-EP-APPL.REAS-DIR-PRF-REAL=REL name.of.village-LOC
‘HER my dad took [as a wife], that’s why she lived all her life in Marankiari.
3.179 He, irotaki intaani ipesakero tonkariki imperetaki.
he irotaki iNtaani i=pes-ak-i=ro
yes FOC only
3m.A=hack.with.a.machete-PRF-REAL=3n.m.O
toNkari-ki iNpereta-ki
top-LOC
rock -LOC
‘Yes, he cleaned ONLY THE ROCKY PART at the top.’
RELATIVIZED OBLIQUE (CAUSE)
3.180 Machiyenka inatzi, irotaki ishipatonatantari.
MachiyeNka i=na-tz-i
irotaki
Matsigenka 3m.S=be-ep-REAL FOC
i=shipatona-t-aNt-a=ri
3m.S=beard-APPL.REAS-REAL=REL
‘He is Machigenka, THIS IS [THE REASON] WHY he has whiskers.’
3.181 Irotaki aisatzi akisavakantapintari.
irotaki aisatzi a=kis-av-ak-aNt-apiNt-a=ri
FOC also
1PL.S=be.angry-RCP-PRF-APPL.REAS-HAB-REAL=REL
‘THIS IS [THE REASON] WHY we argue between ourselves.’
OVERTLY EXPRESSED OBJECT; RELATIVIZED OBJECT
3.182 Irotaki ampeshi niraki, irotaki niriri, ipaitakinari.
irotaki aNpeshi n=ir-ak-i
irotaki n=ir-i=ri
FOC cotton
1SG.S=drink-PRF-REAL FOC 1SG.S=drink-REAL=REL
i=p-ai-t-ak-i=na=ri
3m.A=give-IMP.P-EP-REAL=1SG.O=REL
‘THIS, cotton leaves (tea) I drank, THIS IS WHAT I drank, WHAT they gave me.’
In addition, the verbalized pronouns can also function to focus the referent of an
overtly expressed oblique NP or a copula complement. Note that gender agreement rules
165
are observed. When the pronouns focus an oblique NP or a copula complement, they
immediately precede it, as seen in (3.183-3.184).
3.183 Nayeri, naantari kapichiini kiriiki iroka irotaki notyoonkaantsi.
n=a-ye=ri
n=a-aNt-a=ri
1SG.A=get-IRR=3m.O 1SG.A=get-INST-REAL=3m.O
kapichiini kiriiki
iroka irotaki no=tyooNk-a-ntsi
money DEM FOC 1SG.poss=bathe-EP-NMZ
‘I will get, I earn some money from this TREATMENT.’
little
3.184 Ikantzi acharinite ikinkitsatakotziri aparoni ashaninka kantatziri iritaki
kovintsari.
i=kaNt-tz-i
a=charine-te
3m.S=say-EP-REAL 1PL.poss=grandfather-poss
i=kiNkitsa-t-ako-tz-i=ri
aparoni a=shaniNka
3m.S=tell-EP-APPL-EP-REAL=REL one 1PL.poss=fellowman
kaNt-atsi=ri
iritaki kovintsa-ri
be-STAT=REL
FOC hunt-NMZ
‘They say our ancestors were telling a story about our fellowman who was a
HUNTER.’
Set 4. Additionally, the fourth set of extant pronouns, naintsi, avintsi, irintsi, irontsi,
formed by adding the suffix –intsi to the reduced forms of topic pronouns, is used to
express contrastive topic. In terms of their occurrence in natural discourse, consultants
have commented on the exceedingly rare use of these pronominal forms, indicating that
they have become archaic. In Ashéninka Perené, an oppositive contrastive topic signals a
contrast between two topical referents, the second topical referent being formally
expressed by one of the pronouns from set 4 (Table 25). An English equivalent of the
oppositive contrastive topic is ‘but X’. Propositions with contrastive topics (CT) are often
accompanied by contrastive focused constituents (CF). The following example contains
166
both a contrastive topic (CT) and a contrastive focused constituent (CF): I saw Mary and
John. SHE (CT) says HELLO (CF), BUT HE (CT) is still ANGRY (CF) at you (Lambreht 1994:
291). In (3.185), when the first speaker informs the addressee about his intent to go to
Lima, the addressee interprets it as an invitation to join and responds negatively, ‘BUT
I(CT) don’t
want to go.’ In (3.186), the speaker directly asks ‘Do YOU (CT) grow COFFEE
(CF)?’ and responds to his own question, ‘BUT YOU (CT) have a wooded lot [you have
TREES (CF)].’
In (3.185-3.186), the contrastive topic is signaled by a higher intonational
contour on the pronouns naintsima and avintsi.
3.185
–Nohatatye Irimashiki. –Ari nonintavita nohatemi, naintsima te noninte.
no=ha-t-aty-e
Irimashi-ki ari no=niNta-vi-t-a
1SG.S=go-EP-PROG-IRR Lima-LOC PP 1SG.S=want-FRUS-EP-REAL
no=ha-t-e=mi
naintsi=ma te
no=niNt-e
1SG.S=go-EP-IRR=CNT.F I.C.F=DUB NEG.REAL 1SG.S=want-IRR
‘I will be going to Lima.’‘ I was thinking about going too... but I don’t want to go
[to Lima].’
3.186
–Aviroka tzimatsi piyahaite pivaniki. Avirokaha tzimatsima? Avintsi antamitaki
pivani.
aviroka tzimatsi pi=kahai-te
pi=vani-ki
aviroka=ha
you
EXIST 2poss=coffee-poss 2poss=plot-LOC you=EMPH
tzimatsi=ma avintsi
aNtami-t-ak-i
pi=vani
EXIST=DUB you.C.F
woods-EP-PRF-REAL 2poss=plot
‘You have coffee on your land, do you grow any coffee on your land? Actually,
you have a wooded lot.’
In sum, personal pronouns have more restricted morphological properties than nouns.
For example, they don’t take nominal inflectional and derivational morphology such as
locative case marker –ki, classifying morphemes, diminutive –patsaini, -haniki, or
augmentative –ni; neither do they participate in compounding or noun incorporation.
167
However, they can take demonstrative enclitics =ka ‘this’, =ra ‘that’, =nta ‘that over
there’. Like verbs, pronouns can take modal formatives =ma ‘dubitative’ and =ta
‘optative’ (which seemingly acquire pragmatic functions when used with non-verbal
constituents). They also take floating pragmatic enclitics =ha and =ve. In addition, they
exhibit morphological processes of their own. They take the plural/augmentative marker
–ite (which is also found with adjectives and numerals), although alternatively they may
occur with the nominal plural marker –paye, e.g. arokapaye ‘1PL’, avirokapaye ‘2PL’,
iriropaye ‘3PL.masculine’, iroopaye ‘3PL.non-masculine’. A distinctive feature of the
personal pronouns, which helps identify them as a distinctive class, is ‘emphatic’
pronominal morphology such as formative –ri ‘additive listing focus’, -taki ‘contrastive
exhaustive listing focus’, –intsi ‘contrastive topic’. In syntactic terms, pronouns exhibit
noun-like possibilities for modification. Like nouns, they take modifiers including
numerals and nouns, as seen in (3.187-3.188) respectively; they can also be modified by a
relative clause, as seen in (3.189).
3.187 Aviroka apiteroite antatsini ovantsi.
aviroka apite-ro-ite
aNt-atsi=ni
ovantsi
you
two-NMZ-PL work-STAT=REL plot.of.land
‘You two who will work on the land.’
3.188 Kooyapaye iroo kotsivaitatsini.
kooya-paye iroo kotsi-vai-t-atsi=ni
woman-PL she cook-DUR-EP-STAT=REL
‘All of them women who will be cooking.’
3.189 Aanakiro iroori paitachari Irina.
a-an-ak-i=ro
iroori pait-ach-a=ri
Irina
take-DIR-PRF-REAL=3n.m.O she be.called-STAT-REAL=REL Irina
‘She took it, she, who is called Elena.’
168
(ii) Possessive pronouns. Independent possessive forms derive from the verb root ashi
‘have’ in a regular paradigm, as shown in Table 26.
TABLE 26. Possessive pronouns
Person/Number/
Gender
1SG
2SG
3SG.masculine
3SG.non-masculine
1PL
2PL
3PL.masculine
3PL.non-masculine
Free personal
pronoun/continuous
topic form
naaka
aviroka
iri(ro)
iroo
arokaite
avirokaite
iriroite
iroite
Bound pronoun
form/person-marking
clitic in A/S function
no=
pi=
i=~ir=
o=
a=
Free possessive
pronominal form
nashi
pashi
irashi
Ø-ashi
Ø-ashi
pashiite
irashiite
Ø-ashiite
Free possessive pronominal forms, derived from the verb root ashi ‘to have’, take the
verbal subject person markers, which coincide with the bound possessive proclitics on the
noun. Plural number is expressed by the plural/augmentative marker -ite. (It should be
noted that the plural possessive forms are elicited. In natural discourse, plural forms
haven’t been attested.) The possessive pronouns show gender distinction in the third
person, irashi ‘his’, ashi ‘her’ (the latter can also indicate ‘our’) and irashiite ‘their’
(masculine), ashiite ‘their’ (non-masculine). To show gender agreement with the
controller nouns which constitute the masculine gender class, the affix ir- is assigned.
Possessive pronouns take zero marking when they agree with the nouns which make up
the default, non-masculine gender.
Though possessive pronouns are free forms, they are typically not used by
themselves. They modify the head to express possession or may be used as full NPs in
the copula complement function. As the examples in (3.190-3.192) show, the possessive
169
pronouns are used to express contrastive focus or demonstratively, usually accompanied
with a deictic gesture.
3.190
–Ninka ashitarori maakinaka? –Ashi kooyaka.
niNka ashi-t-a=ro=ri
maakina=ka ashi kooya=ka
who have-EP-REAL=3n.m.O=REL machine=Q hers woman=DEM
‘Whose machine (recorder) is this?’ ‘Hers, this woman’s.’
3.191 Nantyari nomantsaki nonkitsatantyariri irashi neentsipataini.
no=a-aNt-ia=ri
no=mantsaki
1SG.S=get-APPL.REAS-IRR=REL 1SG.poss=cushma
no=N-kitsa-t-aNt-ia=ri=ri
irashi n=eentsi-patsaini
1SG.poss=child-DIM
1SG.A=IRR-dress-EP-APPL.REAS-IRR=3m.O=REL his
‘In order for me to get my cushma and dress my little children.’
3.192 Te kaari nashive!
te
kaari
nashi=ve
NEG.REAL NEG.C mine=EXCL
‘This is not mine.’
Bound pronominal forms. Bound pronouns are phonologically bound clitics which attach
to verbs and nouns to form a phonological word. They have a fixed position with regard
to their hosts. With verbs, the S/A proclitics precede the verb root or, if present, the
irrealis prefix or causative prefix; the O enclitics are verb-final first position clitics.With
noun hosts, the possessive proclitics always occupy the slot before the nominal root.
Table 27 shows forms and functions of the verb and noun proclitics.
TABLE 27. Forms and functions of bound pronominal forms
FORM
n(o)=
p(i)=
i=,y=, ir(i)=
o=
a=
FUNCTION
S/A person marker on verbs
Possessor on nouns
1pSG subject
1pSG possessor
2p subject
2p possessor
3p masculine subject
3p masculine possessor
3p non-masculine subject
3p non-masculine possessor
1pPL subject
1pPL possessor
170
As my corpus shows, free pronouns are used more sparingly than bound pronominal
forms. Free pronominal forms largely correspond to two sets of subject and object person
markers, as seen in Table 28. In Table 28, 1SG and 2SG A/S proclitics show allomorphy.
The vowels of pronominal proclitics are deleted when they are adjoined to vowel initial
stems. The 1PL subject proclitic is realized as a= before non-vowel initial stem or as a
zero morph before vowel-initial stems. As far as the 3SG subjective and objective clitics
are concerned, they are clearly formed from the personal pronouns iri(ro) ‘he’ and iroo
‘she’. Their phonologically reduced forms have retained the gender-based alternation of
/i /and /o/, while the diachronically-motivated loss of the segment /r/ has occurred in the
subjective proclitics. The distribution of the A/S masculine proclitics is as follows: y=
before a-/o- initial stems, Ø before i-initial stems, and i= with the [-vowel]-initial stems.
The non-masculine A/S verbal proclitics have irregularity as well: o= is used before
[-vowel] stem and zero marking occurs before [+vowel] stem.
TABLE 28. Allomorphy of bound pronominal forms
Person/Number/Gender
1SG
2
3SGmasculine
3SGnon-masculine
1PL
Free forms
naaka
aviroka
iri(ro)
iroo
aroka
A/S clitics
n(o)=
p(i)=
i=; y=; Ø; (ir(i)=)
o=; Ø
a=; Ø
O clitics
=na
=mi
=ri
=ro
=ai
Some speakers of the westerly variety of Ashéninka Perené employ the 3SGm subject
marker ir(i)= cumulating with the category of irrealis before vowel-initial or /j/-initial
roots. The irrealis prefix N is realized as an allomorph r- between the 3SGm subject
171
proclitic i= and a vowel-initial root and as ri- between the 3SGm i= and the /j/-initial
verbal root. Examples are provided in (3.193-3.194).
3.193 Te iramitya isaiki kirinka, piyaha.
te
ir=amit-ia
i=saik-e
kiriNka
NEG.REAL 3SG.m.IRR=be.accustomed-IRR 3m.S=be.at-IRR downriver
piy-ah-a
return-REGR-IRR
‘They weren’t accustomed to living downriver, and they came back.’
3.194 Ontzimi aparoni escuela iriyotantahyari noshaninkapaye.
o=N-tzim-e
aparoni escuela
3n.m.S=IRR-have-IRR one
school
iri=yo-t-aNt-ah-ia=ri
no=shaniNka-paye
3SG.m.IRR=learn-EP-APPL.REAS-REGR-IRR=REL 1SG.poss=fellowman-PL
‘A school is needed so that our people learn [how to read and write].’
Allomorphy is also observed with the 3SGm possessor proclitics on nouns: ir=~y=
before a/o- initial stems, Ø before i-/e-initial stems, and i= with the non-vowel initial
stems; e.g. ir=oya-te [3m.poss=pot-poss] ‘his pot’, ir=oki [3m.poss=eye] ‘his eye’,
ir=aiki [3m.poss=tooth] ‘his tooth’, ir=ako [3m.poss=hand] ‘his hand’, ir=otsitzi-te
[3m.poss=dog-poss] ‘his dog’, y=ashiro-tsa-te [3m.poss=iron-CL:long.flexible.thin-poss]
‘his ax’, i=mapi-ni [3m.poss=stone-poss].
In general, person/possesssor proclitics (see Table 27) are obligatory. However,
3pSG intransitive subjects may not be marked on the verb at all. This phenomenon of
zero marking of subject function has a pragmatic basis involving topically continuous
participants (see Payne &Payne 2005 for discussion of split intransitivity in Ashéninka
varieties). Examples are provided in (3.195-3.196).
172
3.195 –Tsika ohatzika ina? –Pokapahi.
tsika o=ha-tz-i=ka
ina
pok-ap-ah-i=Ø
WH 3n.m.S=go-EP-REAL=Q mother come-DIR-REGR-REAL=3n.m.O
‘Where is mother?’ ‘She is back already.’
3.196 Inyapakiri, hataki.
i=ny-ap-ak-i=ri
ha-t-ak-i=Ø
3m.A=see-DIR-PRF-REAL=3m.O go-EP-PRF-REAL=3m.O
‘He saw him and went [towards him].’
As the result of zero subject marking, neutralization of the categories of person,
gender, and number occurs. Another case of neutralization involves the pronominal third
person clitic =ni, which is unmarked for gender. The 3SG clitic =ni occurs when there
are two 3 person participants: one of them is a semantic recipient or a beneficiary
(marked by =ni) and another is a semantic theme or a patient (marked by =ri- or =ro).
Examples are provided in (3.197-3.198).
3.197 Ovarite ironyaaka intaani yayiniro tzina.
o=va-ri-te
ironyaaka iNtaani y=ay-e=ni=ro
tzina
3n.m=eat-NMZ-poss now
solely 3m.A=get-IRR=3O=3n.m.O camonilla
‘Her food, he gave her solely [the pulp of] camonilla.’
3.198 Iposakiri, iminkitakiniri pyaamini itsiyakiki.
i=pos-ak-i=ri
i=mink-it-ak-i=ni=ri
3m.A=beat-PRF-REAL=3m.O 3m.A=insert-ICPL-PRF-REAL=3O=3m.O
pyaameni i=tsiyaki-ki
bow
3m.poss=anus-LOC
‘They beat him and inserted the bow in his anus.’
Additionally, neutralization of the category of reality status occurs when the 1PL
object person marker =ai is used, as seen in (3.199).
173
3.199 Ipakai kiriikiranki.
i=p-ak=ai
kiriiki=raNki
3m.A=give-PRF=1PL.O
money=ADV.P
‘He gave us money [which we don’t have any more].’
(iii) Demonstratives. Demonstratives are defined ‘as a class of shifters with deictic
reference to some person (or something) other than speaker or addressee’ (Dixon
2010a:224). There have been attested three types of demonstratives in the language.
(1) Nominal demonstratives, which occur as free forms in an NP with a noun, e.g. oka
kireeki ‘this money’, iyoka kashiri ‘this monkey’, or as bound forms =ka, =ra, =Nta on
nouns; e.g. maakina=ka [machine=DEM], kooya=ra [woman=DEM] ‘that woman’,
tzivi=Nta [salt=DEM] ‘that salt over there’ (both free and bound forms may be used
together, e.g. iroka kooyaka ‘this woman’.
(2) Locative adverbial demonstratives which occur as free forms, e.g. (h)aka ‘here’, aNta
‘there’, yoNta ‘over there’, or as bound forms =ka, =ra, =Nta on verbs, as seen in
(3.200).
3.200
Nosaikahetzika.
no=saik-a-he-tz-i=ka
1SG.S=be.at-EP-PL-EP-REAL=ADV
‘We are here.’
(3) Locative-existential adverbial gender-sensitive demonstratives hiri-/hiro- ‘here you
are’ which take endings =ka, =ra, =Nto while making a spatial reference to the proffered
entity, as exemplified in (3.201).
3.201 Hironto pivankonta.
hiro=Nto
pi=paNko=Nta
here.it.is=ADV 2poss=house=DEM
‘Here is your house over there, make use of it’.
174
Since adverbial demonstratives have been already discussed in §3.4.4, I’ll provide
here a brief account of nominal demonstrative pronouns. Nominal demonstratives are
organized in two gender-based paradigmatic sets expressing a three-way contrast in terms
of spatial distance from the speaker, as seen in Table 29. Nominal demonstratives
indicate the relative distance of a noun referent vis-à-vis the deictic center which is the
speaker’s location at the time of the utterance. Demonstrative adjectives in Table 29 are
formed by affixing the proximal =ka, medial =ra and distal =Nta to the phonologically
reduced stems iyo-/iri- and iro- of the third person singular personal pronouns iri(ro) ‘he’
and iroo ‘she’.
TABLE 29. Nominal demonstrative pronouns
Gender
masculine
non-masculine
Proximal
(i)irika~ iyoka
iroka~ roka~ oka
Medial
(i)yora
irora
Distal
(i)yoNta
iroNta
Defining properties of nominal demonstratives are having a deictic function and
occurring in an NP with a noun (Dixon 2010a: 227). Nominal demonstratives have a
deictic function in (3.202-3.203) or are used as a way to express new information in
(3.203), while occuring in an NP with a noun in (3.202-3.203) or a head of NP in (3.203).
In (3.204), the nominal demonstrative yoka ‘this’ is used anaphorically. They may appear
as a head of NP in object function, as seen in (3.203), or subject function, as shown in
(3.204).
3.202 Paanakero iroka nihaaka.
p=a-an-ak-e=ro
iroka
2A=take-DIR-PRF-IRR=3n.m.O DEM.n.m.
‘Take this water.’
nihaa=ka
water=DEM
175
3.203 –Pinintatzi irora koka? Iroka pinintziri?
pi=niNt-atz-i
irora
koka iroka
pi=niNt-tz-i=ri
2S=want-PROG-REAL DEM.n.m. coca DEM.n.m. 2S=want-EP-REAL=REL
‘Do you want that coca? Do you want this?’
3.204 Yoka ikovi yayero contrato.
yoka
i=kov-i
y=ay-e=ro
DEM.m 3m.S=want-REAL 3m.A=get-IRR=3n.m.O
‘This [guy] wanted to sign a contract.’
contrato
contract
When nominal demonstratives appear with the corresponding noun, they are often
accompanied with a pointing gesture. Demonstratives such as iroka~ oka ‘this’, irora
‘that’ can also be used for discourse organization as fillers to fill up gaps in utterances
(note that the filler este ‘this’ is frequent in Spanish). Examples are provided in (3.2053.206).
3.205 Tekatsi ante irora ankamitantantyari.
tekatsi
aNt-e
irora
a=N-kamitaNt-aNt-ia=ri
NEG.EXIST do-IRR DEM.n.m 1PL.S=IRR-buy-APPL.REAS-IRR=REL
‘I have nothing to do (filler) to buy food [other than perform steam baths].’
3.206 Oka ironyaaka nosaitatyero nihaa.
oka
ironyaaka no=sai-t-aty-e=ro
nihaa
DEM.n.m. now
1SG.A=empty-EP-PROG-IRR=3n.m.O water
‘This, now I will pour water.’
(iv) Interrogatives. The set of interrogative forms includes only two basic forms such as
tsika ‘where, what, how, why’ and niNka ‘who’. This minimal set of question words
takes advantage of the available periphrastic stratiegies to produce interrogative
constructions by using various copula verbs and verbal affixes and clitics. The form and
morphological patterning of these forms don’t resemble personal or possessive pronouns.
176
The origin of these question words is obscure25. They don’t inflect except for the
dubitative =ma and conditional =rika. Question words are always clause-initial and their
use typically requires the question clitic = ka which attaches to the verb. The question
word tsika is used in quieries about the complements of the copula of naming pait and
temporal, consequence, manner, and locative adjuncts; the question word ninka
interrogates subjects and objects. Interrogative tsika-constructions will be described first,
followed by an account of niNka-constructions.
Tsika. The interrogative pronoun tsika ‘where’ is used in questions about the referent’s
location. Examples in (3.207-3.209) illustrate this function.
3.207 -Tsika noshinto?-Taan, tsikataima.
tsika no=shiNto
taan
tsika=taima
where 1SG.poss=daughter NEG.W where=DUB
‘Where is my daughter?’ ‘I don’t know where she might be.’
3.208 -Tsika norirote? Ari novanakiri inkaranki. Tema hirika.
tsika no=riro-te
ari
no=v-an-ak-i=ri
where 1SG.poss=watch-poss there 1SG.S=put-DIR-PRF-REAL=3m.O
iNkaraNki tema
hirika
recently maybe here.it.is
‘Where is my watch? I put it there a while ago. Maybe, it is here.’
3.209 -Tsikarika naye kompiroshi?-Tzimatsi novanikinta osheki kompiroshi.
tsika=rika
n=ay-e
koNpiroshi tzimatsi
where=COND 1SG.S=take-IRR yarina
EXIST
no=vani-ki=Nta
osheki koNpiroshi
1SG.poss=land-LOC=DEM many yarina
‘Where will I get yarina?’ ‘I have a lot of yarina on my land over there.’
25
Cysouw notes that niNka ‘who’ is probably the reduction of Asháninka janika ja-ni=ka [who-ANIM=Q]
‘who’; presumably, niNka consists of the animate marker -ni and the general interrogative clitic =ka.
Regarding the morphology of tsika tsi=ka [?=Q], the evolution of the root tsi is not clear. The root might be
related to the unspecified possessor suffix –(n)tsi (2007:150-1).
177
Additonally, tsika is used as an argument with the meaning ‘what’ to ask about a
referent’s name and about subject’s actions, as shown in (3.210-3.212). The semantic
roles of the interrogated complements include patients and themes. In general, whatconstructions are expressed by combining tsika with the copula verb of naming pai and
the relative enclitic =ri, marked on the lexical verb, as seen in (3.211-3.213).
3.210 -Tsika opaita(ka) kooyaka?- Te onkameetsate irora, kamaaro onatzi.
tsika o=pai-t-a=ka
kooya=ka
te
what 3n.m.S=be.called-EP-REAL=Q woman=DEM NEG.REAL
o=N-kameetsat-e
irora kamaaro o=na-tz-i
3n.m.S=IRR-be.good-IRR DEM devil
3n.m.S=be-REAL
‘What’s the name of this woman?’ ‘She isn’t good, she’s the devil.’
3.211
-(Tsika) paita yaminiri notsaratiki?-Taantya.
tsika paita y=amin-i=ri
no=tsara-te-ki
taan=tya
WH
3m.S=look.for-REAL=REL 1SG.poss=bag-poss-LOC W.NEG=EMPH
‘What were they looking for in my bag?’ ‘I don’t know.’
3.212
–(Tsika) paita pikantakeriri oimitsori?-Nosampitakotatziri noshintotsori.
tsika paita pi=kaNt-ak-i=ri=ri
o=imi-tsori
WH
2S=say-PRF-REAL=3m.O=REL 3n.m.poss=husband-DIST.REL
no=sampi-t-ako-t-atz-i=ri
no=shiNto-tsori
1SG.A=ask-EP-APPL-EP-PROG-REAL=3m.O 1SG.poss=daughter-DIST.REL
‘What did you say to your cousin?’ ‘I asked him about my niece.’
3.213 Tsikama opaitaka pinintziri?
tsika=ma o=pai-t-a=ka
pi=niNt-tz-i=ri
WH=DUB 3n.m.S=be.called-EP-REAL=Q 2S=want-EP-REAL=REL
‘What do you want?’
As the above examples show, the formulaic expression tsika pai-t-a [where be.called-EPREAL] ‘what is it called that’ has become grammaticalized to signal the complementfocused function of the interrogative form. It has been desemantacizing and losing
phonetic substance to the extent that tsika is often omitted. The copula verb of naming
178
pai ‘be called’ also shows loss of substance since it often appears without the subject
person marker. The same grammaticalization process is underway in why-interrogative
constructions, which rarely occur with the interrogarive form tsika. The elements of the
why-constructions include the copula verb of naming pai, applicative of reason –aNt and
relative clitic =ri, as seen in (3.214).
3.214 -Paitakya? Paitaka ikamantakari otsitzi? –Te, te tekira inkami.
pai-t-a=kya
pai-t-a=ka
i=kam-aNt-ak-a=ri
be.called-EP-REAL=EMPH be.called-EP-REAL=Q 3m.S=die-APPL.REAS-PRF-REAL=REL
otsitzi te
te
tekira i=N-kam-e
dog NEG.REAL NEG.REAL yet
3m.S=IRR-die-IRR
‘Why? Why did the dog die?’ ‘No, it hasn’t died yet.’
Tsika is also used in the questions about the manner of the action performed by the
subject. These questions literally interrogate about the causes of the action. When used in
this function, tsika occurs with the verb kaNt ‘happen’, ‘do’ in conjunction with the
applicative of reason –aNt and relative clitic =ri or conditional =rika, as seen in (3.2153.216).
3.215 -Tsika okanta pinyavaitantakari?-Nonyavaitantari anyaaniki nohataki escuela.
tsika o=kaNt-a
pi=nya-vai-t-aNt-ak-a=ri
WH 3n.m.S=happen-REAL 2S=speak-DUR-EP-APPL.REAS-PRF-REAL=REL
no=nya-vai-t-aNt-a=ri
a=nyaa-ni-ki
1SG.S=speak-DUR-EP-APPL.REAS-REAL=REL 1PL.poss=word-poss-LOC
no=ha-t-ak-i
escuela
1SG.S=go-EP-PRF-REAL school
‘How did it happen that you speak the language [why do you speak]?’ ‘I speak
[the language] because I went to school.’
179
3.216 -Tsika pikantzirorika pichakopitepaye? -Nokintantaro tsimiri.
tsika pi=kaNt-tz-i=ro=rika
pi=chakopi-te-paye
WH 2A=do-EP-REAL=3n.m.O=COND 2poss=arrow-poss-PL
no=kiNt-aNt-a=ro
tsimiri
1SG.A=pierce.with.arrows-APPL.REAS-REAL=3n.m.O bird
‘How do you use your arrows [what for do you use your arrows]?’ ‘To kill birds.’
When questions about quantity (how much? how many?) are asked, tsika is generally
used along with the copula of capacity kara ‘be complete’, ‘contain’ and the relative
clitic =ri. When temporal adverbial adjuncts are interrogated, the relative clitic =ri does
not co-occur with tsika, as shown in (3.217-3.220).
3.217 -Tsika okarate kireeki pimpinatenari? -Osheki.
tsika o=kara-t-e
kireeki pi=N-pina-t-e=na=ri
osheki
WH 3n.m.S=CAP.C-EP-IRR money 2A=IRR-pay-EP-IRR=1SG.O=REL many
‘How much money will you pay me?’ ‘A lot.’
3.218 -Tsika okaratzi nihaa pinintziri? -Tekatsi.
tsika o=kara-tz-i
nihaa pi=niNt-tz-i=ri
tekatsi
WH 3n.m.S=CAP.C-EP-REAL water 2S=want-EP-REAL=REL none
‘How much water do you need?’ ‘None.’
3.219 -Tsika okaratzi pimpokahe? –Shetenitaima.
tsika o=kara-tz-i
pi=N-pok-ah-e
sheteni=taima
WH 3n.m.S=CAP.C-EP-REAL 2S=IRR-come-REGR-IRR afternoon=DUB
‘When will you come back?’ ‘In the afternoon.’
3.220 -Tsika okarateka antavaite?-Ikitaitzimataki.
tsika o=kara-t-e=ka
aNt-a-vai-t-e
WH 3n.m.S=CAP.C=IRR=Q work-EP-DUR-EP-IRR
i=kitai-tz-ima-t-ak-e
3m.S=be.morning-EP-INCH-EP-PRF-IRR
‘How much time will we work?’ ‘Till dawn.’
The expression tsika ikantaitziroka ‘how do they say it?’ (or its shorter version
without tsika) is also found as a modifier of nouns in a situation when speakers express
180
hesitation while looking for the right word. This expression is often preceded by the filler
irora ‘this’. Examples are provided in (3.221-3.222).
3.221 Isaikavitaka aka irora tsika ikantaitziroka nirinive.
i=saik-a-vit-ak-a
aka irora tsika
3m.S=be.at-EP-FRUS-PRF-REAL here DEM WH
i=kaNt-ai-tz-i=ro=ka
n=iri-ni=ve
3m.A=say-IMP.P-EP-REAL=3n.m.O=Q 1SG.poss=father-deceased=EXCL
‘There lived this, whachamacallit, my deceased father.’
3.222 Irora ikantaitzirora imperetatzira.
irora tsika i=kaNt-ai-tz-i=ro=ra
iNpereta-tz-i=ra
DEM WH 3m.A=say-IMP.P-EP-REAL=3n.m.O=ADV rock-EP-REAL=ADV
‘This, what do they call it, the rocky area there.’
The form tsika(rika) may function as a conjunction occurring in a headless relative
clause, as seen in (3.223-3.224).
3.223 Onyi tsika onkameetsate osaiki.
o=ny-i
tsika o=N-kameetsa-t-e
o=saik-e
3n.m.S=look.for-REAL where 3n.m.S=IRR-be.good-EP-IRR 3n.m.S=be.at-IRR
‘She was looking (for a place) where it was suitable to settle down.’
3.224 Pihate tsika pinintakaro.
pi=ha-t-e
tsika pi=niNt-ak-a=ro
2S=go-EP-IRR where 2S=want-PRF-REAL=3n.m.O
‘Go (to a place) where you want to.’
NiNka. The question word niNka ‘who’ often occurs as an argument in questions about
the referent’s identity (who is this?) and possessor (whose?). Examples (3.225-3.226)
illustrate questions about subject and possessor arguments respectively.
181
3.225 -Ninka kooyanta?-Irotaki osheninka noina, irotaki yomitantatsiri.
niNka kooya=Nta
irotaki o=sheniNka
no=ina
irotaki
who woman=DEM FOC 3n.m.poss=fellowman 1SG.poss=wife FOC
yomit-aNt-atsi=ri
teach-APPL.REAS-STAT=REL
‘Who is that woman?’ ‘She is my wife’s compatriot, a teacher.’
3.226 -Ninkama ashitarori iroka?-Ashitaima Irina.
niNka=ma ashi-t-a=ro=ri
iroka ashi=taima Irina
who=DUB have-EP-REAL=3n.m.O=REL DEM hers=DUB Irina
‘Who owns this?’ ‘Maybe [it’s] hers, Elena’s.’
NiNka ‘who/whom’ is used in direct questions about the subject or object arguments.
Examples are provided in (3.227-3.228).
3.227 -Ninka pokatsirika?-Impokatye satantsinkari.
niNka pok-atsi=ri=ka
i=N-pok-aty-e
who come-STAT-REAL=REL=Q 3m.S=IRR-come-PROG-IRR
sat-aNt-tsiNkari
stick-APPL.REAS-ADJ
‘Who is coming?’ ‘A doctor will be coming.’
3.228
-Ninkama okaimiri? -Okaimatziri oyariri.
niNka=ma o=kaim-i=ri
o=kaim-atz-i=ri
who=DUB 3n.m.A =call-REAL=REL 3n.m.A=call-PROG-REAL=3n.m.O
o=yari-ri
3n.m.poss=brother-poss
‘Whom did she call?’ ‘She called her brother.’
(v) Indefinite forms. Indefinite animate forms are represented by niNkarika ‘everybody’,
‘somebody’, maaroni ‘everybody/all’, aparoni ‘somebody’, and specific indefinite
aparopaye ‘each’. These pronouns are derived from the forms whose original functions
have been extended to include an indefinite sense. The indefinite forms are generally
182
positioned before or after the verb. A summary of the indefinite proforms is provided in
Table 30.
TABLE 30. Indefinite forms
Form
niNkarika [niNka=rika]
aparoni
Gloss
‘everybody’
‘somebody’
‘all’,
‘everyone’
‘somebody’
indefinite pronoun
aparopaye [aparo=paye]
‘each’
specific indefinite
maaroni
Function
indefinite pronoun
Other functions
indefinite pronoun
adjective ‘entire’;
quantifier ‘all’
numeral ‘one’; indefinite
article
NiNkarika, which consists of niNka ‘who’ and the conditional clitic =rika, occurs as
an object argument in (3.229) and as a modifier of the subject argument expressed by the
indefinite pronoun maaroni ‘all’ in (3.230).
3.229
Arika inkimaki ninkarika kantzimanintakiriri iritaki intsatsinkaitakeri.
arika i=N-kim-ak-e
niNka=rika
when 3m.S=hear-PRF-IRR who=COND
kantzima-niNt-ak-i=ri=ri
iritaki
speak.badly-DIM.DEGR-PRF-REAL=3m.O=REL FOC
i=N-tsatsiNk-ai-t-ak-e=ri
3m.A=IRR-behead-IMP.P-PRF-IRR=3m.O
‘When he hears somebody slightly badmouth him, they [his henchmen] will
behead this person.’
3.230 Maaroni ikantanakina iyoka, ninkarika tsikarika nohate.
maaroni i=kaNt-an-ak-i=na
iyoka niNka=rika
all
3m.A=call-DIR-PRF-REAL=1SG.O DEM who=COND
tsika=rika
no=ha-t-e
where=COND 1SG.S=go-EP-IRR
‘Everyone calls me this, everybody, wherever I go.’
183
The form maaroni ‘everybody’ occurs as a subject or object argument, as seen in
(3.231-3.232). The plural/augmentative infix <ite> can be inserted for emphasis.
3.231 Maaroetini ishiyahetanaka.
maaro<ite>ni
i=shiy-a-he-t-an-ak-a
everybody<AUG>everybody 3m.S=run-EP-PL-EP-DIR-PRF-REAL
‘Everybody ran.’
3.232 Yapotohetakiri maaroni.
y=apoto-he-t-ak-i=ri
maaroni
3m.A=gather-PL-EP-PRF-REAL=3m.O everybody
‘He gathered everybody.’
The difference between niNkarika ‘everybody’, maaroni ‘everybody’ and aparopaye
‘each’ lies in the degree of specificity of the referent. Only aparopaye ‘each’ is used with
the specific referent in mind. Examples (3.233-3.234) illustrate the use of the general
indefinite form aparoni ‘somebody’ and the specific indefinite aparopaye
‘each/everyone’.
3.233 Ivavanitari irosatzi impokantakya aparoni, amakirori nyaantsi.
i=pavani-t-a=ri
irosatzi i=N-pok-aNt-ak-ia
3m.A=idolize-EP-REAL=3m.O until 3m.S=IRR-come-APPL.REAS-PRF-IRR
aparoni
am-ak-i=ro=ri
nyaa-ntsi
somebody bring-PRF-REAL=3n.m.O=REL word-ABS
‘They had idolized it until somebody came who brought religion.’
3.234 Ari okantatya itsaronkakini ikantahetaki aparopaye.
ari o=kaNt-aty-a
i=tsaroNka-ki-ni
PP 3n.m.S=happen-PROG-REAL 3m.poss=fear-LOC-DIM
i=kaNt-a-he-t-ak-i
aparo-paye
3m.S=be-EP-PL-EP-PRF-REAL one-PL
‘It was this way that each of them was living in fear.’
184
3.4.6 Numerals
The tiny closed class of numerals includes three basic cardinal numbers such as aparoni
‘one’, apite ‘two’, and mava ‘three’. There are also other elicited numerals which exhibit
a great deal of variation. Some speakers insist that the only authentic numbers are the first
three cardinal numbers; the rest has been made up by the local educational authorities. In
general, numerals bigger than ‘three’ do not occur in natural discourse. Instead, Spanish
loans are used, especially, when referring to money. In the past, as speakers point out,
quantifiers such as osheki ‘many’ and pashini ‘[some] others’ were used for numbers
above ‘three’.There are only two ordinal numbers derived from the corresponding
cardinal numbers, e.g. aparonitanaintsiri ‘the one which is first’ and apitetatsiri ‘the one
which is second’; cf. mava kamenaantsi [three advice] ‘the third piece of advice’. The
elicited numeral forms are reproduced in Table 31. The numbers consistently attested in
the elicited data are given in bold.
Numerals as a class can only be identified on semantic grounds. Their morphological
patterning is too heterogeneous to serve as a definitional criterion. Numerals function as
noun modifiers and generally precede the head noun in an NP, as seen in (3.235-3.236).
3.235 Ipokantakari apite kishiisatzi ikantapakiri aparoni ashaninka.
i=pok-aNt-ak-a=ri
apite kishii-satzi
3m.S=come-APPL.REAS-PRF-REAL=REL two pajonalino-provenance
i=kaNt-ap-ak-i=ri
aparoni a=shaniNka
3m.A=say-DIR-PRF-REAL=3m.O one
1PL.poss=fellowman
‘When two pajonalinos arrived, they said to a fellowman.’
185
3.236 Yookaitahi dieciocho, saikanahi siete, cuatro shirampari, mava kooya.
y=ook-a-it-ah-i
dieciocho o=saik-an-ah-i
3m.S=leave-EP-ICPL-REGR-REAL eighteen 3n.m.S=be.at-DIR-REGR-REAL
siete cuatro shiraNpari mava kooya
seven four man
three woman
‘Eighteen left, seven remained, four men and three women.’
TABLE 31. Elicited numeral forms
Gloss
One
Two
Three
Four
Five
Six
Seven
Eight
Nine
Ten
Twenty
Thirty
Forty
Fifty
Sixty
Seventy
Eighty
Ninety
A hundred
Speaker 1 (Bajo
Marankiari)
aparoni
apite
mava
apiteta
apapokoroni
oihatzirori
yaantsiri
ocho
pashini
apitevakaye
apiteniri
mavaniri
apitetaniri
apapokoronini
oihatzironiri
yantsiniri
ochoniri
pashiniri
osheki
Form
Speaker 2 (Pampa
Michi)
aparoni
apite
mava
apite(ka)vakaye
apapakoroni
shintapakerori
shirinkapaichari
pasini
yaatapakirori
oshikiri
oshikiriri
Speaker 3 (San Miguel)
aparoni~apani
apite
mava
otsi
koni
iko
tson
tsoti
tin
tsa
oshikiriri oshikini
3.4.7 Interjections, ideophones, and discourse particles
Interjections and ideophones form a group of phonologically and morphosyntactically
idiosyncratic units which often fall ouside of the phonological system of the language and
its morphological and syntactic processes. Discourse particles do not exhibit
distinctiveness in their phonology. I will discuss each group in detail below.
186
(i) Interjections. An interjection is defined as “a conventionalized cry, typically
indicating the speaker’s emotional response to something that has happened to them, or
something that they have observed or become aware of” (Dixon 2010a:27). In Ashéninka
Perené, interjections, except akaya ‘ouch!’, are not associated with any formatives. They
are monosyllabic (except akaya ‘ouch!’) and therefore don’t follow phonological
requirements for the minimal word which must have two syllables at minimum (see
2.4.3). Interjections stand on their own in the clause and tend to appear on the clause
periphery. The most frequently used interjections are listed in Table 32.
TABLE 32. Interjections
Form
akaya
ha
aa
oo
Gloss
‘ouch!’
‘aha’
‘humph’
‘oh dear’
Function
to express pain, hurt
to express understanding, recognition
to express disbelief, dissatisfaction
to express grief
The intensity of the emotion may be expressed by lengthening of the final vowel and
adding the exclamative clitic =ve, as seen in (3.237-3.239).
3.237 Nokaimanaki akayaveee!
no=kaim-an-ak-i
1SG.S=call-DIR-PRF-REAL
‘I screamed in pain, ‘Ouch!’
akaya=ve-e-e
ouch!=EXCL
3.238 Ha, aka niroka, aa, kaari, apaatara, notomi!
ha
aka n=iroka
aa
kaari
apaata=ra
no=tomi
aha here 1SG.poss=DEM humph NEG.C wait!=ADV 1SG.poss=son
‘Aha, here it is, humph, it isn’t, wait, son!’
187
3.239 -Nopishitziro novanko, notsonka notsamaitayetahiro. –Ha, iro pantaitziri
sheteniranki.
no=pishi-tz-i=ro
no=paNko
no=tsoNk-a
1SG.A=sweep-EP-REAL=3n.m.O 1SG.poss=house 1SG.S=finish-REAL
no=tsamai-t-a-ye-t-ah-i=ro
ha iro
1SG.A=cut.grass-EP-REP-DIST-EP-REGR-REAL=3n.m.O aha this
p=aNta-it-tz-i=ri
sheteni=raNki
2S=work-ICPL-EP-REAL=REL day=ADV.P
‘I swept the house and cut the grass.’ ‘Aha, this is what you did yesterday.’
(ii) Ideophones. By definition, ideophones or sound symbolic expressions, which have a
conventionalized meaning, represent an expressive part of the language to give the
listener a more vivid account of an action, event, or state (Lang 2004:410). As a word
class, ideophones are invariant words which take no morphology. They don’t exactly
follow phonological rules of the language, e.g. although consonantal codas are
disallowed, they are found in the the ideophones tsipak ‘jumping and bouncing off the
surface’, tsinik ‘grab’, taNkorek ‘sudden freezing and then fast disappearance ’. Most of
them are multisyllabic and are pronounced with a relatively high pitch and intensity.
Ideophones are syntactically independent units which can replace a clause. In general,
they appear on the clause periphery but also can occur after the verb or noun with which
they are associated. A list of ideophones (which is not exhaustive by all means) is
provided in Table 33. Repetion of the ideophone expresses repeated actions being carried
out. The number of repetitions depends on the action being carried out; those ideophones
which express the suddenness of the action are not repeated. Word-final vowel in some
ideophones can be lengthened to highlight the intensity of the action, as seen in (3.2403.243).
188
TABLE 33. Ideophones
Form
poo
(tsa)pok
motek
pote
tsikiri
tsimpak
tzinik
tirek
taNkorek
shiparek
takik
mokoro
shero
teriririri
Meaning
hitting something; penetration
falling into water
bouncing off the surface (a round object, e.g. a ball)
slowly trickling drops of liquid
liquid gushing out in a stream
sound of convulsions of the agonizing human body
grab; stop
sound of heavy load falling on the surface
sudden freezing and then fast disappearance
fast disappearance
biting nails
vomiting
slipping and falling on the surface
thunder
3.240 Oposakina noitoki poo-o, notapiki poo-o!
o=pos-ak-i=na
no=ito-ki
poo-o
3n.m.A=hit-PRF-REAL=1SG.O 1SG.poss=head-LOC IDEO
no=tapi-ki
poo-o
1SG.poss=back-LOC IDEO
‘It hit me on the head, bam, on the back, bam!’
3.241 Oshitovanaki osheki nirani, tsikiri, tsikiri.
o=shitov-an-ak-i
osheki n=ira-ni
tsikiri tsikiri
3n.m.S=escape-DIR-PRF-REAL much 1SG.poss=blood-poss IDEO IDEO
‘My blood gushed in a stream.’
3.242 Ikimavakera ikamantakoitziri mitanaka, tankorek.
i=kim-av-ak-i=ra
i=kamaNt-ako-it-tz-i=ri
3m.S=hear-DIR-PRF-REAL=ADV 3m.S=inform-APPL-ICPL-EP-REAL=REL
mi-t-an-ak-a
taNkorek
jump-EP-DIR-PRF-REAL IDEO
‘When he heard the report, he froze and then quickly disappeared.’
3.243 Inoshikakeri tzinik, tzinik, tzinik, okaakitaki konoya.
i=noshik-ak-i=ri
tzinik tzinik tzinik okaaki-t-ak-i
konoya
3m.A=pull-PRF-REAL=3m.O IDEO IDEO IDEO be.close-EP-PRF-REAL turtle
‘He (the boa) pulled him (the turtle) (grabbing sound) and the turtle moved close
to him.’
189
(iii) Discourse particles. Discourse particles such as greetings, calls to attract attention,
short responses to a question yes/no can be considered within the class of interjections
(e.g. Dixon classifies them as such (2010a:29)) or can be treated as a separate class. The
fact that they are conventionalized signs may serve as grounds for including them in the
class of interjections. However, unlike interjections, they do not represent a spontanenous
emotional response of the speaker to an action, event, or state. Neither do they exhibit
distinctiveness in their phonology. Here discourse particles are treated as a separate class.
As shown in Table 34, there are two types of discourse particles: the first type
represents clause-bound utterances which are a conversational turn on the part of the
addressee; the second type includes stand-alone free utterances. Stand-alone free
utterances are semantically, phonologically and grammatically independent from the rest
of the clause which they may precede or which may follow them. Examples are provided
in (3.244-3.247).
TABLE 34. Discourse particles
Syntactic type
Clause-bound
utterance (obligatorily
associated with the
preceding clause)
Stand-alone free
utterance
Form
he
te
ari(ve)
arima
taan(tya)
naakave
tsame
paasoNki
aviro(ka)
hatanara
intsi(tya)
Meaning
yes
no
OK
maybe
have no idea, no clue
hello (in response)
come on
thank you
hello
I am going (good bye)
so…(new topic)
190
3.244 –Noyatanakiro pirento. –Arive, kameetsataki.
n=oya-t-an-ak-i=ro
pi=reNto
ari=ve
1SG.A=follow-EP-DIR-PRF-REAL=3n.m.O 2poss=sister PP=EXCL
kameetsa-t-ak-i
be.good-EP-PRF-REAL
‘I followed your sister.’ ‘OK, that’s good.’
3.245 -Tsika noshinto? -Taantya tsikataima.
tsika no=shiNto
taan=tya
tsika=taima
WH 1SG.poss=daughter W.NEG=N.C.FOC WH=DUB
‘Where is my daughter?’ ‘No idea where she could be.’
3.247 Nintavakyarita oovero; intsitya, oovero.
n=iNt-av-ak-ia=ri=ta
oovero
intsitya oovero
1SG.A=begin-DIR-PRF-IRR=3m.O=OPT hawk.sp. well
hawk.sp.
‘I will begin with the hawk; so, the hawk.’
3.248 Tsame ahate novankoki.
tsame
a=ha-t-e
no=paNko-ki
come.on 1PL.S=go-EP-IRR 1SG.poss=house-LOC
‘Come on, let’s go to my house.’
191
Chapter 4
Clause linking
This chapter deals with clause linkage, examining the ways clauses can be combined in
Ashéninka Perené. An overview of clause linking is provided in Table 35 which uses the
following abbreviations: MC stands for ‘main clause’, SC for ‘subordinate clause’, MC1
for ‘1st main clause’, and MC2 for ‘2nd main clause’. As seen in Table 35, clauses are
linked either via subordination (when a subordinate clause is embedded in the main
clause) or coordination (when two clauses are juxtaposed). The chapter discusses relative
clauses in §4.1, complementation in §4.2, adverbial clauses in §4.3, and other types of
clause linkage in §4.4.
4.1 Relative clauses
Syntactic typology of relative clauses distinguishes between headed, headless, adjoined,
and correlative clauses (Andrews 2007:206-263). When the nominal head occurs in the
main clause, the relative clause is said to be externally headed; when the head NP is
stated in the relative clause, it is categorized as internally headed; when the head noun is
stated in neither, the relative clause is called headless. Arguably, these discreet syntactic
categories do not specify other syntactic subtypes of headed relative clauses with reduced
nominal heads such as demonstrative, indefinite, and negative polarity pronouns; neither
are non-nominal or light main clause heads such as class terms, classifiers, plural number
markers, and pronominal person clitics are recognized (Citko 2004:97-98; Epps 2009;
Dixon 2010a:337-8). To capture variability of heads in Ashéninka Perené relative
clauses, a more finely-grained syntactic typology of relatives is considered here.
192
TABLE 35. Summary of clause linkage
Type of clause linking
Relative, externally
headed
Relative, headless
Non-canonical relative
clause constructions
Complementation
Temporal succession
Temporal simultaneity
(brief overlap)
Temporal simultaneity
(prolonged overlap)
Temporal anteriority
Possible conditional
Counterfactual
conditional
Cause linking
Purposive linking
Motion purpose
Resultative linking
Undesirable possible
consequence linking
Addition
Contrast
Disjunction
Manner
Form
=ri, =ni
Locus of linking
SC verb
Clause order
MC SC
tsika/tsikarika ‘wherever’
tekatsi, tzimatsi
impersonal passive –ai;
apposition
apposition
ironyaaka, i/oponya;
apposition
arika ‘when’;
=ra ‘when’
SC-initial
MC
MC2 verb
MC SC
ovakera ‘when’;
i/okaNta ‘in the meantime’
irohatzi ‘until’;
tekira ‘before’
aririka ‘if’;
=rika ‘if’;
=mi
tema/kama ‘because’;
i/okaNta ‘because’;
apposition
-aNt…=ri ‘in order to’;
oNkaNtya + -aNt..=ri ‘so
that’;
oNkaNtya ‘so that’
-ashi ‘with the intent to’
apposition
irotaki + -aNt..=ri ‘that’s
why’
=kari
apposition
iro/irokya/iroma ‘but’
=ha
kantzimai(n)tacha
terika ‘if not’
kimi ‘be similar’; shiy ‘be
like’
MC2-initial
MC1 MC2
MC1 MC2
MC MC
MC1 MC2
SC-initial
SC verb
SC MC
free
SC-initial
SC-initial
SC-initial
SC-initial
SC-initial
SC verb
MC1 verb,
MC2verb
SC-initial
SC-initial
SC MC
SC MC
MC SC
SC MC
SC MC
SC MC
MC1 MC2
SC verb
SC
SC
MC verb
Free
SC MC
MC MC
MC SC
MC SC
MC SC
MC SC
SC
MC1 MC2
MC SC
SC verb
MC SC
MC2-initial
MC2 personal
pronoun
MC2 initial
MC2initial
MC2initial
MC1 MC2
MC1 MC2
MC1 MC2
MC1MC2
MC1 MC2
MC1MC2
193
In addition to two major relative clause types, identified in Ashéninka Perené- headed
and headless clauses- a distinct intermediate subtype of headed relative clauses is
proposed, which covers both reduced heads (demonstratives, generic terms, numerals),
and light heads (pronominal person clitics and classifiers occurring as a bound verbal
element, not an independent constituent of a clause). Each relative clause type is formally
expressed via specific relativizing strategies, which will be the focus of this presentation.
I will first provide an outline of the =ri/=ni relativizing strategy in §4.1.1, followed by a
discussion of other available means for constructing relative clauses in §4.1.2-4.1.3.
Additional functions of the relativizing forms =ri/=ni are addressed in §4.1.4.
4.1.1 The =ri/=ni relativizing strategy in headed relative clauses
This discussion of Ashéninka Perené headed relative clauses draws on the definitional
criteria for a canonical relative clause proposed by Dixon (2010a:314-315). Specifically,
main clause and relative clause form a single intonational unit and make up one sentence.
The common argument, shared by the two clauses, may be formally stated only in the
main clause or it may be stated in both clauses; however, it is understood to function as
an argument in both main and relative clauses.
Relative clause functions as a syntactic modifier to the common argument in the main
clause. Canonical relative clause in Ashéninka Perené is externally headed, i.e. it follows
the nominal head, stated in the main clause26 . At the semantic level, relative clause either
provides information to specify or focus the unidentified common argument (restrictive
26
While relative clauses follow the NP that they modify, other noun modifiers such as non-derived
adjectives, quantifiers, numerals, and demonstratives precede the head. Nonetheless, the postnominal
pattern is observed in noun compounding N+N when the head of the nominal compound precedes the
modifier, and in derived adjectives which follow the head nominal.
194
relative clause), or provides background information about the already identified
common argument (non-restrictive relative clause). Segmental marking of the headed
relative clause is performed by the relative clause clitics =ri or =ni which go into the slot
of the pronominal verbal enclitic in O function in the transitive relative clause, replacing
it; when the relative clause is intransitive, the relative marker =ri or=ni appears at the
end of the verb. When subjects are relativized, the subject proclitic on the relative clause
verb has zero realization. The default relativizer =ri marks both realis and irrealis events
while =ni is tendentially associated with irrealis events27. Relative clause has a basic
structure of a clause including the predicate and its arguments.
The default relativizing strategy in Ashéninka Perené involves the forms =ri or=ni,
cliticized to the end of the verb. These forms have other functions in Ashéninka Perené
which will be discussed in §4.1.4. As mentioned above, the default relativizer =ri appears
with relative clauses marked for either realis or irrealis status. The functions of the
common argument in the irrealis and realis relative clauses marked by the relativizer =ri
include object of the transitive clause in (4.1- 4.2)28, subject of the transitive clause in
(4.3), and object of the transitive clause in (4.4). Crucially, the =ri strategy does not
apply to common arguments in A or S functions in irrealis relative clauses29.
27
Payne states that the distribution of =ri and =ni depends on the mode of the relative clause, i.e. when the
relative clause refers to the realis event which did happen, the relativizer =ri is used; when the relative
clause refers to an irrealis act which didn’t happen, the relative clause marker =ni is selected (1989:172-4).
However, this observation doesn’t hold for the language communities of Pampa Michi and Bajo Marankiari
which served as the research base for this project.
28
As a result of applicative derivation, the common argument in the relative clause in (4.1) functions as the
verb argument and occupies the direct object slot. In other words, the instrumental sampi ‘torch’ must be
promoted to object and then relativized as a syntactic object. Note that the verb-final relativizer =ri
replaces the bound pronoun in O function; cf. yootantyariro y=oot-aNt-ia=ri=ro [3m.A=shine.light.at-INSIRR=3m.O=3n.m.O] ‘He will shine light at the river with it.’
29
Throughout this section, in my examples I will give the common argument and the relative clause
markers in bold in both the second line containing the segmented text and in the interlinear morpheme-by-
195
4.1
Ihatzi pareniki, yaanaki sampi yootatantyariri nihaaki yameni shima.
[i=hatz-i
pareni-ki
y=a-an-ak-i
saNpi]MC
3m.S=go-REAL river-LOC 3m.S=take-DIR-PRF-REAL torch
[y=oot-a-t-aNt-ia=ri=ri
nihaa-ki
y=amen-e
shima]RC
3m.A=light-REP-EP-INST-IRR=3m.O=REL water-LOC 3m.A=see-IRR fish.sp.
‘He would go to the river and take a torch with which he would shine light at the
river to look for boquichicos (Prochilodis sp.).’
4.2
Airorika ironyaaka nosaatantavaitzi, tekatsi nayeri noyari.
airo=rika
ironyaaka no=saa-t-aNt-a-vai-tz-i
NEG.IRR=COND now
1SG.S=bathe.in.hot.water-EP-CUST-DUR-EP-REAL
[tekatsi ]MC [n=ay-e=ri
no=ya=ri]RC
NEG.EXIST 1SG.A=get.IRR-IRR=REL 1SG.S=eat.IRR=REL
‘If I don’t perform this steam-bathing procedure, there will be nothing that I will
get to eat.’
4.3
Maatsi aka aparoni irora pyatsatakamiri.
[tzimatsi aka aparoni]MC irora
[piyatsa-t-ak-a=mi=ri]RC
EXIST here one
whachamacallit disobey-EP-PRF-REAL=2O=REL
‘There was one who disobeyed you.’
4.4
Ovanitavahiri vatsatsi yamahiri oime.
[o=v-anii-t-av-ah-i=ri
vatsatsi]MC
3n.m.S=CAUS-go-EP-DIR-REGR-REAL=3m.O meat
[y=am-ah-i=ri
o=ime]RC
3m.S=bring-REGR-REAL=REL
3n.m.poss=husband
‘The wife unloaded the meat which had been brought by her husband.’
In contrast, the relativizer =ni tendentially appears with irrealis relative clauses which
relativize subjects of intransitive and transitive verbs, as seen in (4.5-4.7).
morpheme gloss, and use square brackets [] to indicate the boundaries of the clause, along with the
abbreviations MC=main clause, RC=relative clause, CC1=first coordinate clause, CC2=second coordinate
clause.
196
4.5
Nominkimaityaro inchakii, apite okanta, oka saikatsini aka, oka saikatsini anta.
no=miNk-imai-t-ia=ro
inchakii apite
o=kaNt-a
1SG.A=put-INCH-EP-IRR=3n.m.O stick
two
3n.m.S=be-REAL
aka]RC1 [oka] MC2 [saik-atsi=ni
aNta]RC2
[oka]MC1 [saik-atsi=ni
DEM
be.at-STAT=REL here
DEM
be.at-STAT=REL there
‘I will insert this wood stick, there are two wood sticks, this [one] that will stay
here, and this [one] that will stay there.’
4.6
-Ninka ovakahainika? -Naaka, naaka ovakayamini.
niNka ov-ak-ah=ai=ni=ka
[naaka naaka]MC
who eat-CAUS-REGR=1PL.O=REL=Q I
I
[ov-aka-ia=mi=ni]RC
eat-CAUS-IRR=2O=REL
‘Who’s the one who will treat [feed] us?’ ‘I am the one who will be treating you.’
4.7
Kompatsirintaha iritaki potsoteroni ironyaaka ovoroki.
[koNpatsiri=Nta=ha]MC
[iritaki potsot-e=ro=ni
ironyaaka
second.father=DEM=EMPH FOC paint-IRR=3n.m.O=REL now
o=poro-ki]RC
3n.m.poss=face-LOC
‘Her second father is the one who will paint her face.’
Evidence from (4.1-4.7) suggests that the relativizer =ni requires the common argument
to be in A or S function in the irrealis relative clause whereas the default relative clitic
=ri is used in irrealis relative clauses to mark the common argument in O function. When
the ri-strategy is used, the possible functions of the common argument (CA) in the main
clause are as follows: subject of transitive or intransitive clause (A, S), object (O), copula
complement (CC), possessor (POSS), and oblique (OBL); the possible functions of
common argument in the relative clause are subject (A, S) and object (O). Note that the
ri-strategy is used to relativize only subjects of realis relative clauses (RC), whereas both
objects of realis and irrealis relative clauses can be relativized. The collected corpus
provides tentative evidence that when =ni is used to relativize subjects in irrealis relative
197
clauses, the syntactic functions of CA in the main clause are limited to subject, object and
copula complement. The allowed syntactic functions of the common argument (CA) in
the main (MC) and relative clauses (RC) are summarized in Table 36.
TABLE 36. Syntactic functions of the common argument in MC and RC marked by the relativizers =ri and
=ni
Syntactic functions of CA
MC
A,S, O, CC, POSS,
OBL
=ri –marked RC
A,S (realis only)
O (realis, irrealis)
Syntactic functions of CA
MC
A, S, O, CC
=ni-marked RC
A,S (irrealis)
Restrictive and non-restrictive clauses are essentially formally similar. The restrictive
relative clauses in (4.8-4.10) specify the common argument and are pronounced as one
intonational unit with the main clauses. The relative clause in each example is positioned
immediately after the common argument and is marked by the relative enclitic =ri.
4.8
Tzimatsi ochevo airikachari henokinta.
[airik-acha=ri
henoki=Nta]RC
[tzimatsi o=chevo]MC
EXIST 3n.m.poss=branch hang-STAT=REL top=DEM
‘There was a branch which was hanging high at the top of the tree.’
4.9
Pamine mapi kantamirontakitachari.
[p=amin-e
mapi]MC [kaNt-a-miroNt-a-ki-t-acha=ri] RC
2S=look.for-IRR stone
be-EP-flat-EP-CL:small.round-EP-STAT=REL
‘Look over there for a stone which is flat and round.’
4.10
Tzimatsi isorarotepaye katziyavintariri aminatsiri.
[tzimatsi i=soraro-te-paye ]MC
[katziy-a-viNt-a=ri=ri]RC1
EXIST 3m.poss=soldier-poss-PL stand-EP-BEN-REAL=3m.O=REL
[amin-atsi=ri]RC2
see-STAT=REL
‘Apiinka had his soldiers who were standing guard on his behalf and who were
watching over.’
198
As mentioned above, non-restrictive clauses are formally equivalent to restrictive
clauses. In (4.11-4.13), the common arguments are expressed by the NPs which uniquely
identify their human referents as Apiinka, iriroite ‘they (masculine), and aviroka ‘you’.
The relative clause in each example is positioned immediately after the head nominal and
is marked by the relative enclitic =ri or =ni. A seen in (4.8-4.13), the syntactic functions
of the common argument in the relative clauses are limited to the core functions in both
restrictive and non-restrictive types of relative clauses.
4.11
Ironyaaka shetenika nonkinkitsatakoteri aparoni ashaninka ikantaitziri Apiinka
saikaintsirika aka anampikika.
[ironyaaka sheteni=ka no=N-kiNkitsa-t-ako-t-e=ri
aparoni
now
day=DEM 1SG.A=IRR-tell-EP-APPL-EP-IRR=3m.O one
a=shaniNka
1PL.poss=fellowman
i=kaNt-ai-tz-i=ri
ApiiNka]MC
3m.A=call-IMP.P-EP-REAL=3m.O person’s.name
[saik-aintsi=ri=ka
aka a=naNpi-ki=ka]RC
be.at-STAT.T=REL=ADV here 1PL.poss=village-LOC=DEM
‘Now this afternoon, I am going to tell about our fellowman called Apiinka, who
lived here in our village.’
4.12
Iriroite antavaitatziri novaniki nompinatahetatyeri.
[iriro-ite aNtavai-t-atsi=ri
no=vani-ki]RC
he-PL
work-STAT=REL 1SG.poss=land-LOC
[no=N-pinat-a-he-t-aty-e=ri]MC
1SG.A=IRR-pay-EP-PL-EP-PROG-IRR=3m.O
‘Those who were working on my land, I will pay them.’
4.13
Aviroka hatatsini antamiki, pamini vatsatsi avyari.
[aviroka ha-t-atsi=ni aNtami-ki]RC [p=amin-e
vatsatsi a=v-ia=ri]MC
you
go-EP-STAT=REL forest-LOC 2S=look.for-IRR meat
1PL.A=eat-IRR=3m.O
‘You who will go to the woods, you’ll look for game for us to eat.’
199
When the head is expressed as an NP in the main clause, a few possibilities in its
realization have been attested. In addition to common and proper nouns, a number of
associated with an NP elements such as personal pronouns, generic terms such as aparoni
‘one’, maaroni ‘all’, niNkarika ‘whoever’, numerals, and demonstratives can occur as
heads of relatives. Examples of such reduced expressions of nominal heads are provided
in (4.14-4.16). The head nominal expressed by the generic term niNkarika ‘somebody’ is
seen in (4.14), by the demonstrative oka ‘this’ is seen in (4.15), and by the numeral
apiteroite ‘two (animate entities)’ in (4.16).
4.14
Arika inkimake ninkarika kantzimanintakiriri irotaki intsatsinkaitakeri.
[arika i=N-kim-ak-e
niNka=rika]MC
when 3m.S=IRR-hear-PRF-IRR who=COND
[kantzima-niNt-ak-i=ri=ri ]RC
speak.badly-DIM-PRF-REAL=3m.O=REL
[irotaki
FOC
i=N-tsatsiNk-ai-t-ak-e=ri]MC
3m.A=IRR-behead-IMP.P-EP-PRF-IRR=3m.O
‘When he hears someone who is speaking somewhat badly of him, they [his
henchmen] will behead this person.’
4.15
Osheki otenatzi oka kantavaitayetachari.
[osheki o=tena-tz-i
oka]MC [kaNt-a-vai-t-a-ye-t-acha=ri]RC
much 3n.m.S=weigh-EP-REAL DEM
be-EP-DUR-EP-DIST-EP-STAT=REL
‘This is heavy, this [thing] which has many colors.’
4.16
Kantzimaintacha saikatsi apiteroite, iminkitakinari nashirote.
[kantzimaiNtacha saik-atsi
apite-ro-ite ]MC
nevertheless
be.at-STAT
two-NMZ-PL
[i=miNki-t-ak-i=na=ri
n=ashiro-te]RC
3m.A=put-PRF-REAL=1SG.O=REL 1SG.poss=steel-poss
‘Though two [steel nails] remain, the ones that they inserted in my arm.’
200
As stated above, the nominal head can be realized in the main clause as verbal person
clitics, interpreted here as light heads. In (4.17), the common argument is expressed in
the main clause by the 3SGn.m objective person marker =ro whereas in (4.18), the
common argument in the main clause is stated both as the subjective 3SGn.m person
marker o= and the verbal classifier –ki ‘small, round’. The common argument in (4.19),
realized on the verb by the 3SGn.m. objective person marker =ro, corresponds to the
indefinite pronoun tsika paita ‘what (ever)’ in the relative clause.
4.17
Yamakiro tasorinkantsi, yamakiro ovavisakotahaini.
y=am-ak-i=ro
tasoriNka-ntsi
3m.A=bring-PRF-REAL=3n.m.O blessing-NMZ
[ov-avis-ako-t-ah=ai=ni]RC
[y=am-ak-i=ro]MC
3m.A=bring-PRF-REAL=3n.m.O CAUS-pass-APPL-EP-REGR=1PL.O=REL
‘He brought blessings and brought what will save us.’
4.18
Amokakero, osheki okisokitzi kityonkakitatsiri.
a=mok-ak-e=ro
[osheki
1Pl.A=make.holes-PRF-IRR=3n.m.O much
o=kiso-ki-tz-i ]MC
[kityoNka-ki-t-atsi=ri]RC
3n.m.S=be.hard-CL:small.round-EP-REAL be.red-CL:small.round-EP-STAT=REL
‘I will make the holes [with beads], they [the beads] are very durable, [the ones]
which are red.’
4.19
Mayempiri inatzi, ikantero kooya irorave, intaryavaitero, yasankantavaitero tsika
inintakari irirori.
mayeNpiri i=na-tz-i
i=kaNt-e=ro
kooya
irora=ve
womanizer 3m.S=be-REAL 3m.A=do-IRR=3n.m.O woman
DEM=EXCL
i=N-tary-a-vai-t-e=ro
3m.A=IRR-open.legs-EP-DUR-EP-IRR=3n.m.O
[y=asaNk-aNta-vai-t-e=ro]MC
[tsika paita
3m.A=smell-APPL.REAS-DUR-IRR=3n.m.O WH
i=niNt-ak-a=ri
3m.S=want-PRF-REAL=REL
irirori]RC
he
201
‘He was a womanizer, and he would do to the woman this, he would open her legs
and he would smell whatever [private parts] he wanted.’
Another example of a syntactically light head is the subject proclitic o= on the
capacity copula verb kara ‘be complete’, which can be interpreted as the underlying
quantifier ‘all’ or ‘this amount’. In (4.20-4.21), the 3SGn.m person proclitic o= is
understood as the common argument shared by the two clauses.
4.20
Ari okaratzi ikenkitsatakainari aanini.
[i=keNkitsa-t-ak-a-i=na=ri
[ari o=kara-tz-i ]MC
PP 3n.m.S=CAP.C-EP-REAL 3m.A=tell-EP-PRF-REGR-REAL=1SG.O=REL
aani-ni]RC
grandpa-deceased
‘This is all [the amount] that my late grandpa told me.’
4.21
Teha, nosampitemi tsika okaratzi pikovakotenari.
[te=ha
no=saNpi-t-e=mi
tsika o=kara-tz-i ]MC
NEG=EMPH 1SG.A=ask-EP-IRR=2O WH 3n.m.S=CAP.C-EP-REAL
[pi=kov-ako-t-e=na=ri]RC
2A=want-APPL-EP-IRR=1SG.O=REL
‘No, I’ll ask you about the amount that you’ll charge me.’
Clearly, if categorized as the syntactically light head, the pronominal element o= in
(4.20-4.21) occupies an intermediate place on the headedness continuum, ranging from a
fully expressed head nominal to a headless relative clause which will be described in the
next section.
4.1.2 Relativizing strategies in headless clauses
Apart from the default relativizers =ri and =ni, another strategy for creating a relative
clause is via the relative pronoun tsika ‘where’. Such relative clauses are headless,
202
exemplified by [I saw]MC [what he wanted]RC (Comrie and Smith 1977:14). In Ashéninka
Perené, headless clauses have neither a nominal head in the main clause nor the formal
marking of the relative clause by =ri or =ni . The only relativizing device, the relative
pronoun tsika ‘where’ or its derived form tsikarika ‘wherever’, appears in the relative
clause to serve as the common argument for both main and relative clauses. In (4.224.23), the common argument has zero realization in the main clause. The relative clause
is marked by either tsika ‘where’ in (4.22) or by tsikarika ‘wherever’ in (4.23). The
relative pronouns in these examples have a fairly generic sense and can be paraphrased as
‘a place in which’ or ‘any place in which’ respectively. The pronouns function as the
common argument shared by both main and relative clauses. As seen in (4.22-4.23), the
common argument in the tsika-relative clause is always in the oblique (locative) function.
4.22
Onyi tsika okameetsatzi osaiki.
[o=ny-i]MC
[tsika o=kameetsa-tz-i
o=saik-e]RC
3n.m.S=see-REAL
WH 3n.m.S=be.good-EP-REAL 3n.m.S=be.at-IRR
‘She was looking where [for a place in which] it is nice to stay.’
4.23
Ironyaaka avisakiro oponya amini tsika osaiki anta.
[ironyaaka avis-ak-i=ro
oponya
now
pass-PRF-REAL=3n.m.O meanwhile
amin-i]MC
look.for-REAL
[tsika=rika
o=saik-e
aNta]RC
WH=COND 3n.m.S=be.at-IRR there
‘Now she walked along the river and looked for wherever [for a place in which]
she could stay.’
Note that it is possible for the common argument, stated in the main clause by an NP, to
be followed by the tsika-relative clause, exemplified by (4.24).
203
4.24
Yantashitakina tantotsi tsika nosaiki.
[y=aNt-ashi-t-ak-i=na
taNtotsi]MC
3m.A=make-APPL.INT-EP-PRF-REAL=1SG.O corral
[tsika no=saik-i]RC
WH 1SG.S=be.at-REAL
‘My father made a corral for me where [in which] I stayed.’
Another relativizing device is found with the negative and positive polarity copula
verbs tzimatsi ‘existential’ and tekatsi ‘negative existential’ in the headless main clause.
The underlying approximate interpretation of the copula constructions involves the
indefinite pronoun ‘some’ or ‘something’ and the negative pronouns ‘nothing’ or
‘nobody’. The attested syntactic functions of the common argument in the headless
relative clauses are subject and object, illustrated by (4.25-4.26) respectively. The
relativized subjects in (4.25-4.26) lack formal marking, similar to the relativized subjects
in headed relative clauses, which are not marked on the relative clause verb. The relative
clitics =ri or =ni are attached to the end of the verbs in the headless relative clauses. A
gap in place of the surface realization of the common argument in the main copula
clauses is seen in (4.25-4.26).
4.25
Tekatsi nyashirinkapahaini.
[tekatsi]MC [nyashiriNk-ap-ah=ai=ni ]RC
NEG.EXIST annoy-DIR-REGR=1PL.O=REL
‘There won’t be anybody who will bother us.’
4.26
Maatsi namakiri inkaranki aka.
[tzimatsi]MC [n=am-ak-i=ri
iNkaraNki aka]RC
EXIST
1SG.A=bring-PRF-REAL=REL recently here
‘I have [some leaves] which I brought here recently.’
204
To summarize, relative clauses can be placed along the continuum of headedness,
ranging from fully expressed nominal heads to headless clauses, as seen in Figure 15.
FIGURE 15. Continuum of headedness of RC ranging from fully expressed NPs to reduced NPs (pronoun,
demonstrative, and quantifier) to light heads (clitics) to headless
Full
Reduced
Light
Zero expression
_____________________________________________________________________________
←
→
Noun
Pronoun
Verbal person clitic
Demonstrative
Verbal classifier
Quantifier
Verbal distributive plural marker
4.1.3 Other relativizing strategies
This section surveys additional relativizing strategies which formally differ from the
canonical relative clause construction, as defined in §4.1.1. In addition to the relativizing
strategies discussed in §4.1.1 and §4.1.2, the impersonal passive marker –ai is often
associated with non-canonical relative clause construction. In such constructions, the
common argument is realized in the first clause as a full NP, as seen in (4.27, 4.29), or as
an indefinite form, as seen in (4.28), and a co-referential pronominal person clitic in the
second clause. The clause with the impersonal passive construction serves to identify the
indefinite human actor without providing specifics of his/her identity, which can be
recovered by the addressee from the context. The ai-marked constructions are noncanonical in a sense that (i) each coordinated clause can stand on its own; (ii) the
syntactic function of the relativized elements is restricted to subjects; (iii) the relativizing
clitics =ri or =ni are not used in clauses with impersonal passive constructions.
Nonetheless, the semantics of the non-canonical relative clause constructions is
equivalent to that of the headed relative clauses marked by the relativizers =ri or =ni.
205
Language consultants unvariably translate the second clause in a biclausal string with
impersonal passive marking as an embedded relative. The ordering of the clauses is
similar to that of headed relative clauses, formed with the default relativizers =ri or =ni.
The first clause with the fully stated common argument is followed by the second clause,
which typically marks the relativized element with a person clitic, as seen in (4.27-4.29).
4.27
Kamaki, ivyaamini itsiyakiki isataitanakiri.
[kam-ak-i
i=pyaameni
i=tsiyaki-ki]CC1
die-PRF-REAL 3m.poss=bow 3m.poss=anus-LOC
[i=sat-ai-t-an-ak-i=ri]CC2
3m.A=stick-EP-IMP.P-DIR-PRF-REAL=3m.O
‘He was dead, the bow in his anus, which had been inserted [by the two
pajonalinos].’
4.28
Maatsi aka aparoni irora, pyatsatakamiri yamaitakiri.
[[tzimatsi aka aparoni]MC irora
[piyatsa-t-ak-a=mi=ri]RC/C1
EXIST here one
whachamacallit disobey-EP-PRF-REAL=2O=REL
[y=am-ai-t-ak-i=ri] RC2
3m.A=bring-EP-IMP.P-PRF-REAL=3m.O
‘There was here one who disobeyed you, whom they brought [to Apiinka].’
4.29
Ponya oyatanakiro irokave nihaaka ikantaitziro Paukartambo.
[ponya oya-t-an-ak-i=ro
iroka=ve
then
3n.m.A.follow-EP-DIR-PRF-REAL=3n.m.O DEM=EXCL
nihaa=ka]CC1 [i=kaNt-ai-tz-i=ro
Paukartambo]CC2
river=DEM 3m.A=say-IMP.P-REAL=3n.m.O river’s.name
‘She followed this river which is called Paucartambo.’
Apart from the non-canonical relative clauses associated with impersonal passive
constructions, there have been collected other examples of coordinated clauses with the
semantics of relative clauses. In (4.30-4.31), the common argument in the first clause is
expressed by the 3SGn.m.person marker =ro, cliticized to the end of the verb amen ‘see’,
206
and the possessive NP ashi iroori [her she] ‘hers’. The corresponding element in the
second clause is the distributive marker –ye ‘each one’, suffixed to the verbal root
aNpitsa ‘be wrinkled’. As collected data show, the function of the common argument in
the second clause is limited to subjects. Although the default relativizers =ri or =ni are
absent, clause ordering follows the ordering pattern of headed relative clauses; the
common argument in the first clause is modified by the second clause. In (4.31), the
common argument in the first clause (which is also an embedded relative clause) is
realized on the verb pitziry ‘peel’ by the cross-referencing pronominal element =ro
‘3SGn.m.’ and the corresponding proclitic o= ‘3SG.n.m’ and the distributive marker –ye
‘each one’, attached to the verb root vitzi ‘grow hair’ in the second correlative clause.
Note that in the second clause in (4.31), the relativized subject is formally expressed
unlike relativized subjects in headed relative clauses, which are not marked on the
relative clause verb.
4.30
Okantavetakari inkaranki amenapatziro ashi iroori ampitsayetaka.
[o=kaNta-ve-t-ak-a=ri
iNkaraNki
3n.m.S=say-FRUS-EP-PRF-REAL=3n.m.O recently
amen-ap-atz-i=ro
ashi iroori ]CC1 [aNpitsa-ye-t-ak-a]CC2
see-DIR-PROG-REAL=3n.m.O her she
be.wrinkled-EP-DIST-PRF-REAL
‘They [women] said recently that they saw hers [i.e. hand-made bags], which
were crooked.’
4.31
Osaavatashitahiro nihaa, opitziryantaharori ovitzitayetaki akoki.
[o=saava-t-ashi-t-ah-i=ro
nihaa]MC
3n.m.A=heat-EP-APPL.INT-EP-REGR-REAL=3n.m.O water
[o=pitziry-aNt-ah-a=ro=ri ]RC/CC1
3n.m.A=peel-INST-REGR-REAL=3n.m.O=REL
[o=vitzit-a-ye-t-ak-i
3n.m.S=grow.hair-EP-DIST-PRF-REAL
ako-ki]CC2
hand-LOC
207
‘She heated water with which she removed hair which had grown on the girl’s
arms.’
There may be a competing analysis of the second clause in (4.27-4.31) as an apposed
clause, not as a relative clause. However, the apposition analysis would fail on two
counts, namely, semantics and intonation. Examples in (4.27-4.31) demonstrate that it is
impossible to interpret the second clause as an adjacent clause because there is a
conceptual connection between the events described in the clauses. An event in one
clause entails the event in another clause, e.g. She removed the hair which had grown on
the girl’s arms; There was a bow in his anus which had been inserted by two pajonalinos.
The meaning of the above clauses requires a relative clause interpretation. In addition,
there is no pause between the two clauses. They are pronounced as one intonational unit.
In contrast, apposed clauses are marked by a pause and a comma in orthography, when
written down by a literate language consultant.
Yet another way of analyzing the cited data would be to treat the second clause in
examples (4.27-4.31) as the type of right-adjoined clause. However, the criterial
parameters for the adjoined clause do not quite match the Ashéninka Perené noncanonical relative clause constructions. Semantically, adjoined clauses show versatility
and can express temporal, conditional, causal, and purposive types of clausal linking
(Hale 1976:79; Jendraschek 2009:140, 143; Dixon 2010a: 358-359; Andrews 2007: 216217). The ‘correct’ interpretation is often inferred based on contextual clues. In examples
(4.27-4.31), no such ambiguity is observed; only relative interpretation is possible.
Syntactically, adjoined clauses are subordinated and cannot be used by itself to “represent
208
the whole structurally” (Lehmann 2004). In contrast, each clause in examples (4.27-4.31)
can be easily integrated into discourse.
The final type of relative clauses which exhibit non-canonical segmental marking are
te-negated30clauses. Negation limits morphological possibilities for relativizing realis
clauses, thus te-negated relative clauses lack the default relativizer =ri. As shown in
(4.32-4.34), te-negated clauses are distinct in that the affirmative and negative clauses are
coordinated and can stand on their own. They share a common argument which is
expressed in the first clause by the indefinite form aparoni ‘one’ in (4.32), demonstrative
oka ‘this’ in (4.33), and personal pronoun avirokaite ‘you (plural)’ in (4.34). The
relativized elements in the attested negated clauses are subjects. The negated clause
directly follows the head nominal, which is typical of headed relative clause ordering, but
lacks any overt relativizer morpheme. The negated relative clause retains subject marking
on the predicate, contrary to the pattern of zero subject realization of relativized subjects
in headed affirmative relative clauses.
4.32
Aani, novaki aparoni, te ishiyari pashinipaye.
[aani
no=v-ak-i
aparoni ]MC [te
grandseer 1SG.S=kill-PRF-REAL one
NEG.REAL
i=shiy-ia=ri
pashini-paye]RC
3m.S=be.similar-IRR=3m.O other-PL
‘Grandseer, I killed one [bird] which doesn’t look like the others.’
30
There are a few negation strategies in Ashéninka Perené involving the negative particles te (realis) and
airo (irrealis), negative polarity verbs tekatsi ‘negative existential’, kaari ‘negative existential’, ‘be not the
case’, te onkameetsate ‘be not good’, and aritapaki ‘be enough, no more’, and a negative aspectual adverb
tekira ‘not yet’. The default negation strategy is via the negative particles te (realis) and airo (irrealis).
209
4.33
Nokokovavetari oka, osheki te ontenate.
no=kokov~av-a-ve-t-a=ri
1SG.A=want~DIR-EP-FRUS-EP-REAL=3m.O
oka]MC
DEM
[osheki te
o=N-tena-t-e]RC
much
NEG.REAL 3n.m.S=IRR-weigh-EP-REAL
‘I prefer strongly this, which doesn’t weigh much.’
4.34
Avirokaite te pinkimisantahetena ari nompasatakemi.
[ari
[aviroka-ite te
pi=N-kimisaNta-het-e=na ]RC
you-PL
NEG.REAL 2A=IRR-pay.attention-PL-IRR=1SG.O PP
no=N-pasat-ak-e=mi]MC
1SG.A=IRR-punish-PRF-IRR=2O
‘You who didn’t obey me, I’ll punish you.’
In sum, the common argument in the main clause in Ashéninka Perené may be
expressed by a full noun phrase (externally headed clauses) or have zero realization
(headless clauses). The common argument can also be realized by reduced heads
(demonstratives, generic terms, and numerals) and light heads (bound pronominal person
markers and classifiers). Table 37 is a summary of the types of relative clauses.
TABLE 37. Types of relative clauses
HEADED
Statement
of CA
MC/
CC1
RC/
CC2
Relativizer
=ri/=ni
Other relativizing
means
HEADLESS
TYPE 1
HEADLESS
TYPE 2
NONCANONICAL
NP or person
clitic
yes/no
Ø
Ø
NP or person clitic
no
no
person clitic
yes
no
yes
no
copula verbs
tzimatsi ‘EXIST’,
tekatsi
‘NEG.EXIST’ in
MC
impersonal
passive marker–ai
in CC2;
te-negated CC2;
no specific
strategy
relative
pronouns tsika
‘where’/
tsikarika
‘wherever’ in
RC
210
4.1.4 Other functions of the relativizers =ri and =ni
The relative forms =ri and =ni are extremely common and have been found to perform
multiple functions in daily discourse. Their primary function as relativizing forms is to
enable detailed reference to entities. This section focuses on other ways in which the
form =ri can be used such as (i) pronominal marking of 3SGm objects on the verb; (ii)
gender agreement on nouns and non-derived adjectives; (iii) deverbal derivation of
adjectives; (iv) nominalization; (v) a relativizing function with the copula verb of naming
pait, existential copula kant, and copula of capacity kara in content questions; (vi) a
grammaticalized function with the –aNt applicative which has purposive and resultative
senses. The section also covers various functions of the form =ni, which serves to (i)
nominalize verbs; (ii) mark plural number on verbs; (iii) mark alienable possession on
nouns; (iv) form adverbs; (v) produce an intensifying effect when attached to various
hosts. It is hoped that this brief account of multiple functions of the relativizing forms
will contribute to future discussion of the ways in which markers of relative constructions
have evolved. A caveat is due here. Our main goal is to survey the forms and illustrate
their functions with text material. The account of the ri- and ni-forms is presented here
without any attempt to investigate the historical sources of the listed concepts and their
grammaticalization pathways. Some forms are simply homophonous with the relativizing
morphemes =ri and =ni. For example, the alienable possession marker –ri, which is
homophonous with the relativizer =ri, reconstructs to proto-Arawak, most likely as *-re,
whereas the relativizer –ri is an innovation in Kampan languages; or the relativizer =ni is
unlikely to be related to the grammaticalization cline of the class term –ni denoting a
211
deceased referent (Lev Michael, p.c.). Examples of the nominal alienable possessive
suffix –ri and the ‘deceseased referent’ suffix -ni are given in (4.35-4.36) respectively.
4.35
ALIENABLE POSSESSIVE SUFFIX –RI ON NOUNS
Sheteniranki pirento aanahiri oishapaye, ovakari irinchashiri.
sheteni=raNki
pi=reNto
a-an-ah-i=ri
afternoon=ADV.T 2poss=sister take-DIR-REGR-REAL=3m.O
oisha-paye o=v-ak-a=ri
ir=inchashi-ri
sheep-PL 3n.m.A=eat-CAUS-REAL=3m.O 3m.poss=grass-poss
‘Yesterday your sister corraled the sheep to feed them their food [grass].’
4.36
NOMINAL CLASSIFYING ‘DECEASED REFERENT’ TERM -NI
Ari okimita pairani aisatzi naari, chookini ovakera oshitovapahi iraani, irotaki
noyotantarori.
ari o=kimi-t-a
pairani aisatzi naari chooki-ni
PP 3n.m.S=be.similar-EP-REAL before also I
sister-deceased
ovakera o=shitov-ap-ah-i
iraani
irotaki
ago
3n.m.S=leave-DIR-REGR-REAL her.blood FOC
no=yo-t-aNt-a=ro=ri
1SG.A=know-EP-APPL.REAS-REAL=3n.m.O=REL
‘The same happened to me long ago, to my late sister when she menstruated for
the first time, that’s why I know this.’
Functions of the ri-form. As stated above, the ubiquitous form ri performs numerous
functions in everyday discourse. Examples illustrate common uses of =ri as the
pronominal 3m person object enclitic on verbs in (4.37), gender agreement suffix on
nouns in (4.38), gender agreement suffix on adjectives in (4.39), and deverbal
adjectivizing suffix in (4.40).
212
4.37
PERSON OBJECT ENCLITIC ON VERBS
Inyaavakiri irotaki imachokani, kamaki, yamponatakiri maatsirika ivashikaro.
i=ny-av-ak-e=ri
irotaki i=machok-a-ni
3m.A=see-DIR-PRF-REAL=3m.O FOC 3m.S=close.eyes-REAL-AUG
kam-ak-i
y=aNponat-ak-e=ri
tzimatsi=rika i=pashikaro
die-PRF-REAL 3m.A=wrap-PRF-REAL=3m.O EXIST=COND 3m.poss=blanket
‘When they see that he is truly closing his eyes, he is dead, if they have it, they
will wrap him in a blanket.’
4.38
GENDER AGREEMENT SUFFIX ON NOUNS
Aka nosankinatakiri kosaniri.
aka no=saNkina-t-ak-i=ri
kosaniri
here 1SG.A=draw-EP-PRF-REAL=3m.O crocodile.m
‘I’ve painted a crocodile here.’
4.39
GENDER AGREEMENT SUFFIX ON ADJECTIVES
Ponyashitaka ivaitakina antari kitsapi aka noyempitaki.
ponyashitaka i=v-ai-t-ak-i=na
aNtari kitsapi
aka
afterwards
3m.A=put-IMP.P-EP-PRF-REAL=1SG.O big.m needle.m here
no=yeNpita-ki
1SG.poss=ear-LOC
‘Then they stuck a big needle in me here, near my ear.’
4.40
DEVERBAL ADJECTIVIZER
Iroka kipatsi onatzi, kisaarika pitsitsari.
iroka kipatsi o=na-tz-i
kisaa-ri=ka
pitsitsari
DEM mud
3n.m.S=be-EP-REAL be.black-ADJ=DEM clay
‘This is mud, this black clay.’
The nominalizing function of =ri is illustrated by (4.41-4.46). Each nominalization
may occur by itself as an NP, take nominal morphology and modifiers, and/or function as
a complement of complement-taking verbs including a copula complement in copula
clauses. In (4.45), the nominalized verb, which functions as a copula complement, takes
the plural number marker –ite and the modifier apite ‘two’. The nominalized verb hevari
‘the one who leads’, which is in the A function, is apposed to an NP kooya ‘woman’ in
(4.46).
213
4.41
hevatakantzirori
heva-t-ak-aNt-tz-i=ro=ri
lead-EP-APPL.SOC-APPL.REAS-EP-REAL=3n.m.O=REL
‘chief’
4.42
amenakantzirori
amen-ak-aNt-tz-i=ro=ri
see-APPL.SOC-APPL.REAS-EP-REAL=3n.m.O=REL
‘coordinator’
4.43
novantapintari
no=v-aNt-apiNt-a=ri
1SG.S=eat-APPL.REAS-HAB-REAL=REL
‘dining area’
4.44
atsintantapintari
a=tsiNt-aNt-apiNt-a=ri
1PL.S=urinate-APPL.REAS-HAB-REAL=REL
‘toilet’
4.45
Maatsi apite ivantapintarite ivankoki.
tzimatsi apite i=v-aNt-apiNt-a=ri-ite
i=paNko-ki
EXIST two 3m.S=eat-APPL.REAS-HAB-REAL=REL-PL 3m.poss=house-LOC
‘He has two dining areas [the ones they usually use for eating] in his house.’
4.46
Hevari kooya te onintakaina ovitsatyaro onampiki.
heva-a=ri
kooya te
o=niNt-ak-a-e=na
lead-REAL=REL woman NEG.REAL 3n.m.A=want-PRF-REGR-IRR=1SG.O
no=vitsa-t-ia=ro
o=naNpi-ki
1SG.A=visit-EP-IRR=3n.m.O 3n.m.poss=be.in.big.quantity-LOC
‘The female chief [the one who leads] didn’t allow me to visit her community.’
The relativizer =ri also appears on the verbs with the question words tsika ‘where’
and ninka ‘who’ in interrogative constructions involving various copula verbs. Examples
include (4.47-4.49), taken from my transcript of a discussion of the road accident in the
adjacent coffee-growing village of Pichanaki (see §3.4.5 for more examples and
discussion of these constructions).
214
4.47
-Ninka kamaintsiri Pichanakinta? -Kamaki antakotzirori.
niNka kam-aintsi=ri Pichanaki=Nta kam-ak-i
who die-STAT=REL Pichanaki=DEM die-PRF-REAL
aNt-ako-tz-i=ro=ri
work-APPL-EP-REAL=3n.m.O=REL
‘Who died in Pichanaki?’ ‘The driver [the one who managed the car] died.’
4.48
-Tsika okaratzika atziri tetachari kaarokira? -Pashini atziri.
tsika o=kara-tz-i=ka
atziri tet-acha=ri
kaaro-ki=ra
WH 3n.m.=COP.CAP-EP-REAL=Q person fill-STAT=REL car-LOC=DEM
pashini atziri
few
person
‘How many people were in the car?’ ‘A few.’
4.49
-Paitaka oposantakari kaarora? -Antakotzirori te yaminapirotero kametsa,
yaminavaitzi pareniki.
pai-t-a=ka
o=pos-aNt-ak-a=ri
kaaro=ra
be.called-EP-REAL=Q 3n.m.S=crash-APPL.REAS-PRF-REAL=REL car=DEM
aNt-ako-tz-i=ro=ri
te
y=amin-a-piro-t-e=ro
work-APPL-EP-REAL=3n.m.O=REL NEG.REAL 3m.A=see-EP-AUG-EP-IRR=3n.m.O
y=amin-a-vai-tz-i
pareni-ki
3m.S=see-EP-DUR-EP-REAL river-LOC
‘Why did the car crash?’ ‘The driver didn’t pay attention, he was looking at the
river.’
Finally, the relativizing form =ri co-occurs with the instrumental/reason applicative
–aNt to convey a purposive sense. Most likely, the form =ri in the purposive clause
–aNt… =ri ‘in order to’, ‘for’ has grammaticalized from the original
nominalizing/relativizing sense ‘the one who/that/which’ to that with bleached semantics
(see §4.3 for more specifics about this construction). Examples are given in (4.50-4.51).
215
4.50
Iroka pachaka naantari nihaa, ovakerani tekatsi noyate, irotaki iroka
novantarori pyarentsi.
iroka pachaka n=a-aNt-a=ri
nihaa ovakera-ni
DEM pumpkin 1SG.S=take-APPL.REAS-REAL=REL water recently-AUG
tekatsi
n=oya-te
NEG.EXIST 1SG.poss=pot-poss
irotaki iroka
FOC DEM
no=v-aNt-a=ro=ri
pyarentsi
1SG.A=place-APPL.REAS-REAL=3n.m.O=REL manioc.beer
‘This pachaka (pumpkin) I used for carrying water in the old times, there were no
pots, this I (also) used for keeping in it masato (manioc beer).’
4.51
Ontzimatye charine intsikanatai okantyakya okatsitantyari achakopite.
ontzimatye charine
i=N-tsikana-t=ai
o=kaNt-ia=kya
be.necessary grandfather 3m.A=IRR-pepper-EP=1PL.O 3n.m.S=happen-IRR=EMPH
o=katsit-aNt-a=ri
a=chakopi-te
3n.m.A=be.in.pain-APPL.REAS-REAL=REL 1PL.poss=arrow-poss
‘Our grandfather must feed us hot pepper so that our arrows are fatal.’
Functions of the ni-form. The synchronic functions of the form ni show a great deal of
variability. Manifestations of the various functions of the form ni include a deverbal
nominalization, as seen in (4.52), a nominal alienable possession marker, as seen in
(4.53), and a plural number marker on verbs, as seen in (4.54).
4.52
DEVERBAL NOMINALIZER
Yamitakotahetakai amine pashini nyaavaihetatsini.
y=amitako-t-a-he-t-ak=ai
amen-i
pashini
3m.A=help-EP-REP-PL-EP-PRF=1PL.O 1PL.S.look.for-REAL other
nyaavai-he-t-atsi=ni
speak-PL-EP-STAT=REL
‘They helped us look for new language consultants.’
4.53
NOMINAL ALIENABLE POSSESSIVE MARKER
Iroka notyanipaye, iroka noshitzikakero.
iroka no=tya-ni-paye
iroka no=shitzik-ak-i=ro
DEM 1SG.poss=fabric-poss-PL DEM 1SG.A=weave-PRF-REAL=3n.m.O
‘These are my woven products, these I weave.’
216
4.54
VERBAL PLURAL NUMBER MARKER
Sheteni akarahetaiyeni maaroni nosaikahetzika.
sheteni a=kara-he-t-aiy-i-ni
maaroni no=saik-a-he-tz-i=ka
day
1PL=COP.CAP-PL-EP-REAL-PL all
1SG.S=be.at-PL-EP-REAL=ADV
‘Good afternoon to all of us who are here.’
The form ni also serves as an adverbializer, e.g. okaakini ‘closeby’, shintsini
‘quickly’ (see more examples and discussion of adverbs in §3.4.4). When functioning as
an intensifier, ni can attach to various hosts such as verbs, adverbs, and adjectives. The
intensifier may have different interpretations depending on the context. With adjectives,
the form ni intensifies properties of noun referents, e.g. tenari-ni ‘really heavy’, kisori-ni
‘really hard’, osheki-ni ‘really a lot of’. It also has an augmentative sense when it occurs
with adverbs, e.g. ovakera-ni ‘really long ago’. Its attested meanings with verbs include a
habitual meaning ‘one who always acts like this’, as seen in (4.55-4.56); in (4.57), it
intensifies a temporal (remoteness of time) dimension of the carried out action,
essentially expressing an augmentative sense.
4.55
Iraaka, okantziro, ‘Piraakapoo, ari pikantaitatyani pirapiraata.’
ira-ak-a
o=kaNt-tz-i=ro
p=ira-ak-a=poo
ari
cry-PRF-REAL 3n.m.A=say-EP-REAL=3n.m.O 2S=cry-PRF-REAL=EMPH PP
pi=kaNta-it-aty-a-ni
pirapira~t-a
2S=be-ICPL-PROG-REAL-AUG 2S.cry~EP-REAL
‘She was crying, and the mother said to her, ‘You cry too much, you’re always
this way, you cry and cry.’
4.56
Osheki itsarovanitzi, te iyonite imanatya.
osheki i=tsarov-a-ni-tz-i
te
i=yo-ni-t-e
much 3m.S=be.afraid-EP-AUG-EP-REAL NEG.REAL 3m.S=know-AUG-EP-IRR
i=mana-t-ia
3m.S=fight-EP-IRR
‘He has a lot of fear, doesn’t know how to fight.’
217
4.57
Ari ikantakotari mamaro ikoshitzi aparoni kooyaniki irohatzi yatzirivetani.
ari i=kaNt-ako-t-a=ri
mamaro i=koshi-tz-i
PP 3m.S=say-APPL-EP-REAL=REL owl
3m.S=kidnap-EP-REAL
aparoni kooya-aniki
irohatzi y=atziri-ve-t-a-ni
one
woman-DIM
when
3m.S=person-FRUS-EP-REAL-AUG
‘This was the story of the owl which kidnapped a little girl when it was a person
long ago.’
A summary of the synchronic uses of the forms ri and ni is given in Table 38.
TABLE 38. Synchronic functions of the forms ri and ni
ri
3m person object marker on verbs
Gender agreement marker on nouns and
non-derived adjectives
Adjectivizer
Nominalizer
Relativizer
Grammaticalized use with purposive
constructions
ni
Plural marker on verbs
Alienable possessive marker on nouns
Adverbializer
Nominalizer
Relativizer
Intensifier (augmentative)
4.2 Complementation
In §4.1, I discussed relativizing strategies which Ashéninka Perené uses to modify the
common argument shared by the main and dependent clauses. In this section, the focus
will be on the complementation strategy, used to produce a dependent clause which
functions as a verb argument. By definition, a complement clause is a type of clause
which functions as an argument of another clause, has the internal structure of a clause
including at least core arguments, and describes a proposition such as a fact, activity, or
potential (not a place or time) (Noonan 2007:52; Dixon 2010a:370). In this presentation, I
draw on the functional, rather than syntactic, definition of complementation, understood
218
here as expressing a particular relation between events, when one event (the one coded by
the main clause, or the main event) entails another event (the one coded by the
complement clause, or the dependent event). The functional definition of complement
clauses covers the traditional cases of complements (embedded clauses, possibly
introduced by a complementizer) as well as other cases that might not be regarded as
such under traditional syntactic criteria (Cristofaro 2008). In light of this functional
treatment of complementation, the terms ‘main clause’(MC) and ‘complement clause’
(CC) refer here to the ‘main event’ (ME) and ‘dependent event’(DE) respectively.
In Ashéninka Perené, there is no dedicated complement clause construction, i.e. the
language lacks any kind of formal marking whose function is to identify the complement
clause as such, either a distinct complementizer, special marking on the subject or verb of
the complement clause, or stress shift. Instead, the language uses apposition as a
complementation strategy or a grammatical mechanism “for coding a proposition which
is remembered, wanted, seen, heard, known, believed, liked, etc.”(Dixon 2010a:371).
A competing interpretation of Ashéninka Perené complement clause would be an
analysis of its irrealis status marking (which occurs with a limited set of mental state
verbs, e.g. kov ‘want’, nimot ‘like’, amet ‘be accustomed, yomet teach, pintsa ‘decide to
go’, etc.) as a complementation marking strategy (for discussion and examples of special
TAM marking of complement clauses see Dixon (2010a:383-4) and Michael (2007: 15).
Alternatively, the morphosyntactic criteria for examination of complement clause
marking proposed by Cristofaro (2008) state that verb forms in embedded clauses may be
balanced or deranked; a balanced verb is the one that can occur in an independent
declarative clause whereas a deranked verb may lack some of the categorial (TAM)
219
distinctions or display special markers not used in independent clauses. On the basis of
these criteria, Ashéninka Perené complement clauses are regarded as balanced since
complement clauses with mental state verbs can occur as independent declarative clauses.
The appositional complement strategy involves two juxtaposed finite verbs, with the
first verbal predicate functioning as the main clause and the second as the apposed
complement clause. An NP or associated with it element (demonstrative, generic term,
personal pronoun), having reference to the apposed clause, may precede or follow the
apposed verbal predicate. In (a-d), appositional linking is illustrated by the juxtaposed
finite verbs. The verb of volition kov ‘want’ in (a), the thinking verb ooy ‘expect’ in (b),
the phasal verb tsoNk ‘finish/end up’ in (c), and the positive polarity verb ari ‘be the
case’ in (d) are linked to the verbs they relate to within the sentence. Although formally
each clause in the examples in (4.58-4.61) can stand on its own, e.g. ‘he wanted’; ‘she
will respect him, his wife’ in (4.58)‘; ‘I expected it’; ‘it will fall’ in (4.59); ‘she finished’;
‘it grew on her breasts’ in (4.60); ‘this is the case’; ‘he was cutting out his thighs’ in
(4.61), the underlying semantics of the main clause verbs in (4.58-4.61) requires
clarification or specification of what is wanted, expected, or finished.
4.58
Irika evankarika ikovi oninteri iina.
[irika evaNkari=ka i=kov-i]MC
[o=niNt-e=ri
i=ina]DC
this.m young.man=this 3m.S=want-REAL 3n.m.A=love-IRR=3m.O 3m.poss= wife
‘This young man wanted his wife to love him.’
4.59
Nooyamachevetaro omparyanaki.
[n=ooya-macheve-t-a=ro]MC
[o=N-pary-an-ak-e]DC
1SG.A=expect-FRUS-EP-REAL=3n.m.O 3n.m.S=IRR-fall-DIR-PRF-IRR
‘I was expecting it to fall.’
220
4.60
Otsonka ovitzitaki oniiki.
[o=tsoNk-a]MC
[o=vitzit-ak-i
o=nii-ki]DC
3n.m.S=finish-REAL 3n.m.S=grow.hair-PRF-REAL 3n.m.poss=breast-LOC
‘The woman ended up growing hair on her breasts.’
4.61
Ari itovatsatziro ivori.
[ari]MC [i=tova-vatsa-tz-i=ro
i=pori]DC
PP
3m.A=cut-flesh-EP-REAL=3n.m.O 3m.poss=leg
‘That he was cutting out his thighs was the case.’
The grammatical profile of complement clauses is addressed in §4.3.1. Semantic
classes of complement-taking verbs and types of complement clauses they tend to occur
with are discussed in detail in §4.3.2.
4.2.1 The grammatical profile of the complement clause
As mentioned above, the Ashéninka Perené apposed complement clause does not have
distinct segmental marking. The main and complement clauses are basically formally the
same. Specifically, the verb arguments in the complement clause are expressed in the
same way as those in the main clause including the pronominal person clitics. A
complement clause in O function is generally realized by the 3n.m. person enclitic on the
main clause verb.There are no constraints on the expression of peripheral participants in
the complement clause, as exemplified by (4.62-4.63).
4.62
Osheki inevetaro ihate ishimatya pareniki.
[i=ha-t-e
[osheki i=neve-t-a=ro]ME
much 3m.A=like-EP-REAL=3n.m.O 3m.S=go-EP-IRR
i=shima-t-ia
pareni-ki] DE31
3m.S=catch.fish-EP-IRR river-LOC
‘He liked to catch fish in the river.’
31
Examples of complementation clauses are notated in the following way: square brackets [] refer to a
clause, the subscript ME denotes ‘main event’, the subscript DE codes ‘dependent event.’
221
4.63
Airo osamanitzi inyatyero ishirinkavatsavoritanahe.
[airo
osamani-tz-i
i=ny-aty-e=ro]ME
NEG.IRR some.time.later-EP-REAL 3m.A=see-PROG-IRR=3n.m.O
[i=shiriNka-vatsa-pori-t-an-ah-i]DE
3m.S=grow-flesh-thigh-EP-DIR-REGR-REAL
‘It didn’t take long for him to see that his flesh had grown back.’
The available choices of verbal inflectional and derivational formatives are essentially
unrestricted for the complement clause. The complement clause verb exhibits a full set of
affixes and clitics from the attested inventory of inflectional and derivational processes.
Likewise, there are no restrictions on the complement clauses with intransitive subjects,
formally expressed as objects32. Examples of the complement clauses with intransitive
subjects in O function are provided in (4.64-4.65).
4.64
Oshiyakantzi oime pokahi.
[o=shiyakaNt-tz-i]ME
[o=ime
pok-ah-i=Ø]DE
3n.m.S=imagine-EP-REAL 3n.m.poss=husband come-REGR-REAL=3m.O
‘She imagined that her husband had come.’
4.65
Notsarovatzi ashitacha avotsi.
[no=tsarov-atz-i ]ME
[ashi-t-acha=Ø
avotsi]DE
1SG.S=fear-PROG-REAL enclose-EP-STAT=3n.m.O road
‘I am afraid that the road is closed [for traffic].’
However, restrictions apply to the reality status marking. The verbs of volition kov
‘want’, niNt ‘want’, ‘like’, ‘prefer’, ‘respect’, the verb of knowledge yomit~vamet
‘teach’, thinking verbs pintsa ‘decide’, maisaNt ‘forget’, amet~amit ‘be accustomed to’,
ooy ‘expect’, and verbs of liking nimo(t) ‘please’, neve~nive ‘like’ occur only with
complement clauses in irrealis. Note that this group includes mental state verbs which
32
The Payne &Payne review of split intransitivity processes points out that in other varieties of Ashéninka
complement-like clauses prohibit O-inflection on verbs (2005:45).
222
belong to three different semantic domains: desire (‘want’), emotion (‘like’), and
cognition (‘decide’, ‘expect’, ‘forget’, ‘be accustomed to’). Evidently, when mental state
verbs take irrealis complement clauses, this represents the speakers’ view of complement
clause events as hypothetical. In addition, irrealis status restriction involves verbs of
manipulation kov ‘want’, niNt ‘want’, and yomit/vamet ‘teach [=make accustomed to
something]’ whose subjects or agents/manipulators manipulate the behavior of another
participants (objects or manipulees). Irrealis marking with these verbs indicates that
attempted causation may not be successful. In (4.66-4.69), mental state verbs carry an
expectation of a following clause to specify what is wanted, liked, decided, forgotten,
being accustomed to, expected, or taught. Crucially, the action or state of the complement
clause verb is regarded as non-realized when these verbs occur with the main clause
predicates.
4.66
Te pininte paye piimi.
pi=imi]DE
[te
pi=niNt-e]ME [p=ay-e
NEG.REAL 2S=want-IRR 2S=take-IRR 2poss=husband
‘You don’t want to marry your husband.’
4.67
Osheki inevetaro ishimatya tsiteni.
[osheki i=neve-t-a=ro]ME
[i=shima-t-ia
tsiteni]DE
many 3m.A=like-EP-REAL=3n.m.O 3m.S=catch.fish-EP-IRR at.night
‘He liked to fish at night.’
4.68
Pintsatahana nohatahe.
[pintsa-t-ah-a=na]ME
decide-EP-REGR-REAL=1SG.O
‘I am thinking about returning.’
[no=ha-t-ah-e]DE
1SG.S=go-EP-REGR-IRR
223
4.69
Nomaisantakiro nonkaimero noniro.
[no=maisaNt-ak-i=ro] ME
[no=N-kaim-e=ro
1SG.A=forget-PRF-REAL=3n.m.O 1SG.A=IRR=call-IRR=3n.m.O
no=niro]DE
1SG.poss=mother
‘I forgot to call my mother.’
There are also constraints on the co-referential expression of subject in the main and
complement clauses with certain types of complement-taking verbs such as nimot
‘please’ and neve~nive ‘like’, amet~amit ‘be accustomed to’, pintsa ‘decide’, iNt
‘begin’, tsoNk ‘finish’. With other verbs co-referentiality restrictions do not apply.
Examples illustrating the required co-referentiality of the subjects in the main and
complement clauses with the verb of liking nimot and amet ‘be accustomed to’ are given
in (4.70-4.71) respectively.
4.70
Inyashirinkakiro yorave yora shimpokipaye, tsitsiriranki, te onimotero osaiki.
[i=nyashiriNk-ak-i=ro
yora=ve
yora shiNpoki-paye tsitsiri=raNki
3m.A=PRF-REAL=3n.m.O DEM=EXCL DEM horsefly-PL cricket=ADV.P
te
o=nimot-e=ro]ME
[o=saik-e]DE
NEG.REAL 3n.m.S=like-IRR=3n.m.O
3n.m.S=be.at-IRR
‘She was annoyed by the horseflies and crickets, and didn’t like being there.’
4.71
Pairani ashitanitachari osheki yamitaro imposanitya.
[pairani ashi-t-ani-t-acha=ri
osheki y=amit-a=ro]ME
before have-EP-child-EP-STAT=REL much 3m.A=be.accustomed-REAL=3n.m.O
[i=N-pos-ani-t-ia]DE
3m.S=IRR=beat-child-EP-IRR
‘In the old times parents were accustomed to child-beating.’
Negated complement clauses are attested with mental state verbs of thinking, e.g.
shiyakaNt ‘think, imagine’, ny ‘think’, kiNkishiri ‘believe, think’ and emotion, e.g. tsarov
224
‘fear’, and with verbs of speaking, e.g. kaNt ‘say’, kamanaNt ‘inform’ when these verbs
take direct speech complements. Tendentially, factual (realis) complement clauses are
negated. Negated complement clauses pattern the same way as negated main clauses, i.e.
the realis negative particle te requires the clause to be in irrealis. In (4.72-4.75), the
mental states verbs introduce negated direct speech reports, which explains why standard
negation strategies apply to the negated complement clauses. The negative copula verb
kaari generally occurs with non-verbal predicates, as in (4.72-4.73); with standard verbal
predicates, the negative particle te is used, as seen in (4.74-4.75).
4.72
Nokinkishiryaka kaari aviroka.
[no=kiNkishiri-ak-a]ME [kaari
aviroka]DE
1SG.S=think-PRF-REAL NEG.COP you
‘I thought it wasn’t you.’
4.73
Pipyakotakaro kaari pivani.
[pi=pi-ako-t-ak-a=ro]ME
[kaari
2A=lose-APPL-EP-PRF-REAL=3n.m.O NEG.COP
‘You forgot that it isn’t your land.’
4.74
pi=vani]DE
2poss=land
Nonyahantzi te pimpoki.
[no=ny-ah-aNt-tz-i]ME
[te
pi=N-pok-e]DE
1SG.S=see-REGR-APPL.REAS-REAL NEG.REAL 2S=IRR-come-IRR
‘I thought you didn’t come.’
4.75 Notsarovatzi te noyotero inyaani.
[no=tsarov-atz-i]ME
[te
no=yo-t-e=ro
1SG.S=fear-PROG-REAL NEG.REAL 1SG.A=know-EP-IRR-3n.m.O
i=nyaa-ni ]DE
3m.poss=word-poss
‘I’m afraid I don’t know his language.’
The attested syntactic functions of complement clauses are subjects and objects.
Complement clauses fill the S argument slot with the intransitive polarity verbs
225
ari/aritaki ‘be the case’, kaari ‘not be the case’, aritapaki ‘be enough’, evaluative verbs
kameetsataki ‘be good’ and te onkameetsate ‘be bad’. Complement clauses occupy the
default O argument slot with other verbs. In general, O argument slots can be
alternatively filled with an NP.
In terms of their distribution, complement clauses immediately follow the main clause
verb unless the order is reversed for a pragmatic reason. A possible explanation of the
MC CC [main clause-complement clause] positional restriction is that the basic
constituent order in the transitive clause is VOA or AVO. As mentioned above, the
complement clause typically functions as the O argument which occurs after the verb.
The subject argument is often omitted by the speakers on the assumption that the
addressee is able to infer it from the context whereas the object argument tends to be
specified. Nonetheless, an adverb or a core argument may be inserted between the main
clause verb and the complement clause.
4.2.2 Semantic types of complement-taking verbs
Complement-taking verbs include primary verbs and secondary verbs, the former
referring directly to an activity or state and the latter providing semantic modification for
a primary verb to which they are linked either as a grammatical element or as a lexeme
within a complement clause (Dixon 2010a:399). In this presentation of semantic types of
Ashéninka Perené complement-taking verbs, I will draw on Dixon’s typology of
complement-taking verbs (CTV). The lists of complement-taking primary verbs (which is
not exhaustive by all means) and secondary verbs are given below.
226
PRIMARY VERBS:
(1) VOLITION: kov ‘want’, niNt ‘want’, ‘like’, ‘prefer’, ‘respect’;
(2) PERCEPTION: amin ‘see, look for, examine’, ny ‘see’, ‘discover’, ‘think’, kim ‘hear’,
‘listen’, ‘sense’;
(3) LIKING, HABIT: nimot ‘please’, neve~nive ‘like’; amit ~amet ‘be accustomed to’;
(4) KNOWLEDGE, ACQUISITION: yo ‘know, realize, learn’, a ‘grasp’, ‘pick up’, ‘learn’,
yomit~vamet ‘teach’;
(5) FEAR: tsarov ‘have fear’;
(6) THINKING: shiri ‘believe’, shiyakaNt ‘think’, ‘imagine’; kiNkishiri ‘think’,
‘remember’; maisaNt ‘forget’, pi ‘forget’; ooy ‘expect’, ‘wait’; pintsa ‘decide to go’.
(7) SPEAKING: kaNt ‘say’, kamaNt ‘inform’, kaim ‘call out’, exhort’, kamen~kamin
‘advise to prevent something undesirable’, kaNtaNt ‘advise’, kamit ‘plead’, kantzimaniNt
‘threaten’, ‘speak badly’, ak ‘respond’, saNpi ‘ask a question’.
SECONDARY VERBS:
(1) NEGATIVE AND POSITIVE POLARITY INTRANSITIVE VERBS: ari/aritaki ‘be the case’,
kaari ‘be not the case’, aritapaki ‘be enough’; kameetsataki ‘be good/appropriate’, te
oNkameetsate ‘be bad/inappropriate’;
(2) MODAL VERBS: kaNt ‘be able’, ontzimatye ‘must, be necessary’;
(3)PHASAL VERBS: tsoNk ‘finish’, iNt ‘begin’;
(4) MANIPULATIVE VERBS: tsani ‘resist’, shini ‘permit’, amitako ‘help’.
The afore-mentioned classes of verbs tend to co-occur with certain semantic types of
227
complement clauses. At least three types of complement clauses are distinguished based
on the proposition that they describe: (i) the FACT type (refers to the fact that something
took place), e.g. I heard [Mary say goodbye]; (ii) the ACTIVITY type (refers to an activity
extended in time), e.g. He liked [to teach kids]; (iii) the POTENTIAL type (refers to the
possibility for the complement clause subject to be involved in an activity), e.g. I want
[him to go] (Dixon 2010a: 389-392). In what follows, I’ll address the common
occurrences of each verb class with certain types of complement clauses. It is assumed
that the attested preferences of the main clause verb to select specific complement clause
type are motivated by the overlap in the semantic content of the main clause verbs and
the complement clause.
4.2.2.1 Complement-taking primary verbs
There is a large open set of complement-taking primary verbs which make up a complete
sentence through filling their argument slots by NPs, pronouns, clitics, or complement
clauses. As mentioned in §4.2.2, they are divided into seven semantic types whose
profiles are summarized in Table 39.
(1) Volition verbs.The verbs of volition kov ‘to want’, niNt ‘to want’, ‘to like’, ‘to prefer’,
‘to love’ always take a potential complement clause by virtue of their semantics of a
hypothetical, non-realized action. The main clause volition verb in (4.76) occurs with the
frustrative marker -ve which indicates that the attempted action didn’t bring the desired
effect. The complement clause volition verbs often take the counterfactual marker =mi
signaling that the projected action wasn’t realized due to certain circumstances, as
illustrated by (4.76, 4.78-4.79).
228
TABLE 39. Primary complement-taking verbs (CTV)
Semantic
class of
CTV
CTV
form
CTV
gloss
volition
kov
niNt
liking, habit
neve~ nive
nimot
amet~amit
tsarov
‘want’
‘want’, ‘prefer’,
‘like’
‘like’
‘please’
‘be accustomed to’
‘fear’
fear
perception
Semantic type
of CC
Negated
CC?
Subjects
coreferent?
O
O
potential
potential
no
no
no
no
O
O
O
O
potential
potential
potential
potential,
fact
activity, fact
activity
no
no
no
yes
yes
yes
yes
no
no
no
no
no
activity,
fact
fact, activity,
potential
no
no
yes
no
‘look for’, ‘see’
‘see’, ‘think’,
‘find out’
‘listen’, ‘hear’,
‘feel’
‘know’,
‘realize’,
‘learn’
‘teach’
O
O
O
potential
no
no
O
O
potential
fact
yes
yes
no
no
shiyakaNt
maisaNt
pi
pintsa
kaNt
‘believe’, ‘think’
‘remember’,
‘think’
‘imagine’, ‘think’
‘forget’
‘forget’
‘decide to go’
‘say’
O
O
O
O
O
yes
yes
no
no
yes
no
no
no
yes
no
kamen
‘advise against’
O
yes
no
kamet
‘advise in favor’
O
yes
no
ak
‘respond’
O
yes
no
saNpi
‘ask’
O
yes
no
kaim
‘call out’
O
yes
no
kamaNt
‘inform’
O
fact, potential
potential, fact
potential, fact
potential
fact, acitivity,
potential
fact, acitivty,
potential
fact, activity,
potential
fact,
acitivity,
potential
fact, acitivity,
potential
fact, acitivity,
potential
fact, acitivity,
potential
yes
no
amin
ny
kim
knowledge,
acquisition
yo
thinking
yomit~
vamet
shiri
kiNkishiri
speaking
Syntactic
function
of CC
O
O
229
4.76
Okovaveta oshiyapitsatyarimi.
[o=kov-a-ve-t-a=ta]ME
[o=shiy-a-pitsa-t-ia=ri=mi]DE
3n.m.S=want-EP-FRUS-EP-REAL=OPT 3n.m.A=run-EP-APPL.SEP-EP-IRR=3m.O=CNT.F
‘She wanted to run away from him.’
4.77
Te ininte avatsate kapicheeni.
[te
i=niNt-e]ME
[a=vatsa-t-e
NEG.REAL 3m.S=want-IRR
1PL.S=gain.weight-EP-IRR
‘He didn’t want us to eat enough.’
4.78
Nonintzi nonkamemi.
[no=niNt-tz-i ]ME
[no=N-kam-e=mi]DE
1SG.S=want-EP-REAL 1SG.S=IRR-die-IRR=CNT.F
‘I wanted to die.’
4.79
Ikovi itzinanahemi pashini, te inkantahe.
[i=kov-i]ME
[i=tzin-an-ah-e=mi
3m.S=want-REAL 3m.S=rise-DIR-REGR-IRR=CNT.F
kapicheeni ]DE
little
pashini]DE
other
te
i=N-kaNt-ah-e
NEG.REAL 3m.S=IRR-be.able-REGR-IRR
‘Another man wanted to rise [to power], but he wasn’t able [to do that].’
(2) Perception verbs. The perception verbs amin ‘to see’, ‘to look for’, ‘to examine’, ni
‘to see’, to discover’, ‘to find’, and kim ‘to hear’, ‘to listen’, ‘to sense’ tend to take
activity complement clauses, as seen in (4.80-4.82).
4.80
Maaroetini ishiyahetanaka yaminakotahetina tsika okanta noshitovantari.
[maaro<ite>ni i=shiy-a-he-t-an-ak-a
all<PL>all
3m.S=run-EP-PL-EP-DIR-PRF-REAL
[tsika o=kaNt-a
y=amin-ako-t-a-he-t-e=na]ME
3m.S=see-APPL-EP-REP-PL-IRR=1SG.O WH 3n.m.S=happen-REAL
no=shitov-aNt-a=ri]DE
1SG.S=escape-APPL.REAS-REAL=REL
‘Everyone came running to see how I leave the corral.’
230
4.81
Aririka ahate ashimatya nihaaki, anyapakerika ishiyanaki paampatzi.
aririka a=ha-t-e
a=shima-t-ia
nihaa-ki
when 1PL.S=go-EP-IRR 1PL.S=fish-EP-IRR river-LOC
[a=ny-ap-ak-e=rika]ME
[i=shiy-an-ak-e
paaNpatzi ]DE
1PL.S=see-DIR-PRF-IRR=COND 3m.S=run-DIR-PRF-IRR bird.species
‘When we go fishing to the river, we’ll see the bird paampatzi running around.’
4.82
Ikimakeri inyaaventziro potsotzi icharini.
[i=nyaa-veNt-tz-i=ro
[i=kim-ak-e=ri]ME
3m.S=hear-PRF-IRR=3m.O 3m.A=speak-APPL.BEN-EP-REAL=3n.m.O
potsotzi i=charini]DE
annatto 3m.poss=grandfather
‘He heard his grandfather speak about the annatto (Bixa orellana).’
At least two perception verbs kim ‘to hear’, ‘to listen’ and ny ‘to see’, ‘to find’ were
attested to take fact complement clauses, as seen in (4.83-4.84).
4.83
4.84
Nokimaki pikovi itzime peentsite.
[no=kim-ak-i]ME [pi=kov-i
i=tzim-e
1SG.S=hear-PRF 2S=want-REAL 3m.S=have-IRR
‘I heard that you want to have children.’
p=eentsi-te]DE
2poss=child-poss
Opokapahi ina, onyapahina nosaiki intakiro.
o=pok-ap-ah-e
ina
[o=ny-ap-ah-e=na]ME
3n.m=come-DIR-REGR-IRR mother 3n.m.S=see-DIR-REGR-IRR=1SG.O
[no=saik-i
iNtakiro]DE
1SG.S=be.at-REAL outside
‘When my mom would come, she would see me outside the corral.’
(3) Verbs of liking and habit. As mentioned previously, the verbs nimot ‘to please’,
neve~nive ‘to like’, amet ~amit ‘to be accustomed to’ take a potential complement clause,
as seen in (4.85-4.87). A distinction between the formal expression of the experiencer in
(4.85-4.86) reflects the difference in the semantic content of the two verbs of liking. In
(4.85), the experiencer is marked by the subject proclitic on the verb neve ‘to like’
231
whereas the verb nimot ‘to please’ selects the object enclitic for the experiencer function
in (4.86).
4.85
Osheki nonivetaro nonyaatsatya.
[osheki no=nive-t-a=ro]ME
[no=nyaatsa-t-ia]DE
much 1SG.A=like-EP-REAL=3n.m.O 1SG.S=play-EP-IRR
‘I liked a lot to play [soccer].’
4.86
Te onimoteri arika inkatsite imotya.
[te
o=nimot-e=ri]ME
NEG.REAL 3n.m.S=please-IRR=3m.O
[arika
when
i=N-katsit-e
3m.S=IRR-feel.pain-IRR
i=motya]DE
3m.poss=stomach
‘It doesn’t please him when his stomach hurt.’
4.87
Noshaninkapaye yamitaro inkoshite varipa.
[no=shaniNka-paye
y=amit-a=ro]ME
1SG.poss=fellowman-PL 3m.A=be.accustomed-REAL=3n.m.O
[i=N-koshi-t-e
varipa]DE
3m.S=IRR-steal-EP-IRR chicken
‘My fellowmen were accustomed to stealing [freely roaming] chickens.’
(4) Verbs of fear. The verb tsarov fear’ can take either a fact complement clause, as seen
in (4.88), or a potential complement clause, as seen in (4.89).
4.88
Notsarovatzi ashitacha avotsi.
[ashi-t-acha=Ø
avotsi]DE
[no=tsarov-atz-i ]ME
1SG.S=fear-PROG-REAL enclose-EP-STAT=3n.m.O road
‘I am afraid that the road is closed [for traffic].’
4.89
Notsarovanintatzi naniite tsiteni.
[no=tsarov-a-niNt-atz-i]ME
[n=anii-t-e
tsiteni]
1SG.S=be.afraid-EP-DIM-PROG-REAL 1SG.S=walk-EP-IRR at.night
‘I slightly fear to walk in the darkness.’
232
(5) Verbs of knowledge and acquisition. The verb of knowledge yo ‘to know’, ‘to
realize’, ‘to learn’ is attested with all three complement clause types, depending on the
particular sense of the verb. The verb yomit ~vamet ‘to teach’, ‘to train’ generally occurs
with a potential complement clause. In (4.90), when the verb yo has the sense ‘to realize’,
it takes a fact complement clause; in (4.91-4.92), yo ‘to learn’ relates to the activity
clause, in (4.93), the negated verb yo ‘know’ takes a potential complement clause. In
(4.94), the verb vamet ‘be accustomed to’ refers to a potential type clause with which it
generally occurs.
4.90
Iyovetaka manitzira yamatavitaitatziri.
[i=yo-ve-t-ak-a
manitzi=ra]ME
3m.S=know-FRUS-EP-PRF-REAL tiger=DEM
[y=amatavi-t-ai-t-atz-i=ri]DE
3m.A=deceive-EP-IMP.P-EP-PROG-REAL=3m.O
‘The tiger realized that they had deceived him.’
4.91
Okanta ayotantahari kapichiini anyaanavaitzi.
[o=kaNt-a
a=yo-t-aNt-ah-a=ri]ME
3n.m.S=be-REAL 1L.A=learn-EP-APPL.REAS-REGR-REAL=REL
[kapichiini a=nyaana-vai-tz-i] DE
little
1PL.S=read-DUR-EP-REAL
‘This is how we learned to read a little bit.’
4.92
Okanta naari noyotantakari naavintantavaitzi.
[o=kaNt-a
naari no=yo-t-aNt-ak-a=ri]ME
3n.m.S=be-REAL I
1SG.A=learn-EP-APPL.REAS-PRF-REAL=REL
[n=aaviNt-aNt-a-vai-tz-i]DE
1SG.S=cure-CUST-DUR-EP-REAL
‘This is how I learned to cure.’
233
4.93
Te iyote inyaanate, te iyote isankinate.
[te
i=yo-t-e]ME
[i=nyaana-t-e]DE
NEG.REAL 3m.S=know-EP-IRR 3m.S=read-EP-IRR
[te
NEG.REAL
i=yo-t-e]ME
[i=saNkina-t-e]DE
3m.S=know-EP-IRR 3m.S=write-EP-IRR
‘They didn’t know how to read or write.’
4.94
Osheki inevetaro ivametairi evankaripaye tsika onkantya ivayeritya.
[[i=vamet-a-e=ri
[osheki i=neve-t-a=ro]ME
much 3m.A=like-EP-REAL=3n.mO 3m.A=teach-REGR-IRR=3m.O
i=vayeri-t-ia]DE
evaNkari-paye]DE ]ME [tsika o=N-kaNt-ia
young.men-PL
WH 3n.m.S=IRR-do-IRR 3m.S=be.warrior-EP-IRR
‘He liked a lot to teach young people how to fight.’
(6) Verbs of thinking. The thinking verbs shiri ‘to believe’, shiyakaNt ‘to imagine’, ‘to
think’, pintsa ‘to decide to go’, kiNkishiri ‘to think’, ‘to remember’, ny ‘to understand’,
‘to see’, pi ‘to forget’, maisaNt ‘to forget’ select fact or potential complement clauses,
depending on the temporal reference of the direct speech complement they take. In (4.954.98), these mental state verbs function as utterance predicates taking direct speech
complements.
4.95
Naaka noshiritakotaro nariityaro.
[naaka no=shiri-t-ako-t-a=ro]ME
[n=arii-t-ia=ro]DE
I
1SG.S=believe-EP-APPL-EP-REAL=3n.m.O 1SG.S=achieve-EP-IRR=3n.m.O
‘I believed that I would achieve it [I believed, ‘I will achieve it.’]’
4.96
Noshiyakantzi kamakimi.
[no=shiyakaNt-tz-i]ME
[kam-ak-i=mi]DE
3n.m.S=imagine-EP-REAL die-PRF-REAL=2O
‘I imagined that you had died [I thought, ‘You have died’].’
4.97
Nokinkishiryashita tsika pihataki.
[no=kiNkishiri-ashi-t-a]ME
[tsika pi=ha-t-ak-i]DE
1SG.S=think-APPL.INT-EP-REAL WH 2SG=go-EP-PRF-REAL
‘I was wondering where you had gone [I was thinking, ‘Where did you go?’].’
234
4.98
Nonyahantzi aviroka pinatzi.
[no=ny-ah-aNt-tz-i]ME
[aviroka pi=na-tz-i]DE
1SG.S=see-REGR-APPL.REAS-EP-REAL you
2S=be-EP-REAL
‘I presumed it was you [I thought, ‘It is you.’].’
(7) Speaking verbs. Likewise, the verbs of speaking kaNt ‘say’, kamaNt ‘inform’, kaim
‘call out’, exhort’, kamen~kamin ‘advise to prevent something undesirable’, kamit
‘plead’, kantzimaniNt ‘threaten’, ak ‘respond’, saNpi ‘ask a question’ take fact or
potential complement clauses, depending on the complement clause reference time.
Examples are provided in (4.99-4.100).
4.99
Kantzimaintacha charine ikantzi, ‘Airo aviri ivairo koshiri.’
[kantzimaintacha charine i=kaNt-tz-i]ME
[airo
nevertheless
grandpa 3m.S=say-EP-REAL NEG.IRR
a=v-i=ri
i=vairo
koshiri]DE
1PL.A=put-REAL=3m.O 3m.poss=name monkey
‘Nevertheless, my grandpa said, ‘We won’t give him the name ‘monkey’.
4.100 Ikantzi, ‘Tsika isaikika?’ Ikantzi, ‘Hiri isaikinta.’
[i=kaNt-tz-i]ME
[tsika i=saik-i=ka]DE
[i=kaNt-tz=i]ME
[hiri
3m.S=say-EP-REAL WH 3m.S=be.at-REAL=Q 3m.S=say-EP-REAL there
i=saik-i=Nta]DE
3m.S=be.at-REAL=ADV
‘The pajonalino asked [another man], ‘Where is he?’ He answered, ‘He is sitting
there.’
When speaking verbs convey indirect speech, indicated by the change in the deictics,
they take irrealis complement clauses, exemplified by (4.101-4.103).
4.101 Okamitakiri impinateri osheki kireeki.
[o=kamet-ak-i=ri]ME
[i=N-pinat-e=ro
3n.m.A=plead-PRF-REAL=3m.O 3m.A=IRR-pay-IRR=3n.m.O
‘She pleaded with him to pay her more money.’
osheki kireeki]DE
much
money
235
4.102 Ikantzimanintakeri yanteri isheninka.
[i=kaNtzimaniNt-ak-i=ri]ME
[y=aNt-e=ri
3m.A=speak.badly-PRF-REAL=3m.O 3m.A=beat-IRR=3m.O
i=sheniNka]DE
3m.poss=fellowman
‘He threatened to beat up his fellowman.’
4.103 Pairani acharinite ikamenahetzi ari okitaitamanaite.
[ari
[pairani a=charini-te
i=kamen-a-he-tz-i]ME
before 1PL.poss=grandparents-poss 3m.S=advise-EP-PL-EP-REAL PP
o=kitait-aman-a-it-e]DE
3n.m.S=be.morning-early-EP-ICPL-IRR
‘Long ago our grandparents advised [us] on the early hours of the day.’
4.2.2.2 Complement-taking secondary verbs
Complement-taking secondary verbs express ‘secondary’ ideas such as ‘can’, ‘be
necessary’, ‘be the case’, etc. serving to modify the semantic content of the primary verb
(Dixon 2010a: 394-5). As mentioned above, the semantic types of complement-taking
secondary verbs include polarity, modal, phasal, and manipulative verbs. Complements
of all polarity verbs appear in S function whereas complements of other complementtaking secondary verbs occur in O function. It should be noted that the pragmatic
function of the polarity verbs is to intensify the context-determined meanings of
complement clauses without specifying the scale of polarity. Neither positive nor
negative polarity complement-taking verbs mark low or maximal points on the polarity
scale as in ‘be absolutely the case’, or ‘be never the case’; rather, they express a general
clause-level emphasis. Based on the collected corpus, women use significantly more
polarity verbs than men. A summary of the secondary complement-taking verbs is given
in Table 40.
236
TABLE 40. Secondary complement-taking verbs (CTV)
Semantic class
of CTV
CTV
form
CTV
gloss
Polarity
verbs
ari/aritaki
be the case
kaari
not be the case
S
aritapaki
kameetsataki
te
onkameetsate
kaNt
ontzimatye
iNt
tsoNk
shini
tsani
amitako
be enough
be good
be not good
be able
be necessary
begin
finish/end up
allow
resist, prevent
help
Modal
verbs
Phasal verbs
Manipulative
verbs
Syntactic
function
of CC
S
S
S
S
Semantic
type of
CC
fact,activity,
potential
fact,
acitivity,
potential
activity
fact, activity
potential
O
O
O
O
O
O
O
potential
potential
fact, activity
fact, activity
potential
potential
potential
Negated
CC
no
Coreferentiality
restrictions
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
yes
yes
no
no
no
(1) Positive polarity verbs ari/aritaki ‘be the case’. The verbs ari/aritaki ‘be the case’ are
identified as such on the basis of their distributional and structural properties. The verbs
do not occur as verb arguments, are not modified by adjectives or other associated with
nouns elements; do not take nominal morphology such as plural marker -paye,
possessive, diminutive or augmentative morphemes, or nominal classifiers. In contrast,
they function as heads of intransitive predicates and take some verbal morphology (ari is
an invariant verb and aritaki is derived from ari plus the perfective suffix -ak and the
realis status suffix –i).The positive polarity intransitive verb ari ‘be the case’ occurs with
fact, activity, and potential complement clauses in S function, as seen in (4.104-4.106).
4.104 Ari ikantahantana.
[ari]ME [i=kaNt-ah-aNt-a=na]DE
PP
3m.A=say-REGR-APPL.REAS-REAL=1SG.O
‘That they told me [it] is the case.’
237
4.105
Ari itovatsatziro ivori.
i=pori]DE
[ari]ME [i=tova-vatsa-tz-i=ro
PP
3m.A=cut-flesh-EP-REAL=3n.m.O 3m.poss=leg
‘That he was cutting out his thighs was the case.’
4.106 Nokiyakoterita ironyaaka manii, ikinterota, ari oshevatanake.
no=kiy-ako-t-e=ri=ta
ironyaaka manii
1SG.S=poke-APPL-EP-IRR=3m.O=OPT now
bullet.ant
i=kiNt-e=ro=ta
[ari]ME
[o=sheva-t-an-ak-e]DE
3m.A=sting-IRR=3n.m.O=OPT
PP
3n.m.S=move-EP-DIR-PRF-IRR
‘Maybe if I poke the nest of bullet ants, they will probably sting her, that she will
move will be the case.’
The intransitive polarity verb aritaki ari-t-ak-i [PP-EP-PRF-REAL] is regarded by
language consultants as being semantically equivalent to the intransitive positive polarity
verb ari ‘be the case’. The syntactic behavior of aritaki is similar to that of ari: it appears
clause-initially; takes complement clauses in S function; and occurs with all three
semantic types of complement clauses (fact, activity, potential). Examples are provided in
(4.107-4.109).
4.107
Aritaki ikantari, te ayotero avetsikavaheri.
[aritaki]ME [i=kaNt-a=ri]DE
te
a=yo-t-e=ro
PP
3m.S=be-REAL=REL NEG.REAL 1PL.A=know-EP-REAL=3n.m.O
a=vetsik-av-ah-e=ri
1PL.A=fix-DIR-REGR-IRR=3m.O
‘That they are this way is the case, we don’t know how to fix them.’
4.108 He, aritaki impakimi.
he [aritaki]ME [i=N-p-ak-e=mi]DE
yes PP
3m.A=IRR-give-PRF-IRR=2O
‘Yes, that he will give coca to you will be the case.’
238
4.109
Aritaki intaakai pava, irora itasorinka onatzi.
[aritaki]ME [i=N-ta-ak=ai
pava]DE
irora i=tasoriNka
PP
3m.A=IRR-burn-PRF=1PL.O solar.deity DEM 3m.poss=power
o=na-tz-i
3n.m.S=be-EP-REAL
‘That Pava will burn us will be the case, his power is such.’
(2) Negative polarity verb kaari ‘be not the case’. Morphosyntactic properties of the
invariant morpheme kaari are those of the verb; it exhibits verb-like distributional
properties, i.e. it does not function as head of NPs, neither does it occur as verb argument.
The verb always appears as head of intransitive clauses. The negative polarity verb kaari
‘be not the case’33 takes fact complement clause in S function, as seen in (4.110-4.113).
As the examples in (4.110-4.112) show, kaari is regularly used with verbal predicates in
the contexts when the speaker speculates about the cause of the actor’s activity or state.
In this function, kaari appears in conjunction with the applicative of reason/
instrumentality –aNt which has either a resultatve sense, as in (4.110, 4.112) or exhibits
causal semantics, as seen in (4.111).
4.110 Kaari pamantari iitzipaye maniro?
[kaari]ME [p=am-aNt-a=ri
i=itzi-paye
maniro]DE
NEG.P 2A=bring-APPL.REAS-REAL=REL 3m.poss=foot-PL deer
‘Why didn’t you bring the feet of the deer?’ (‘Why that you brought the feet of
the deer wasn’t the case?’)
4.111 Kaari amitantapainta kirinka, piyaha.
[kaari]ME [amit-aNt-apaiNt-a
kiriNka]DE
NEG.P be. accustomed-APPL.REAS-for.a.while-REAL downriver
piy-ah-a
come.back-REGR-REAL
33
With non-verbal predicates, kaari functions as a negative existential copula verb (for discussion of the
copula kaari ‘negative existential’ see §3.2.2).
239
‘Because this wasn’t the case that they were accustomed to living downriver, they
came back.’ (‘They came back because that they were accustomed to living
downriver wasn’t the case.’)
4.112 Irotaki kaari itzimanta iiriikite ipinatahiri yatziritepaye.
irotaki [kaari]ME [i=tzim-aNt-a
i=iriiki-te
FOC
NEG.P 3m.S=have-APPL.REAS-REAL 3m.poss=poney-poss
i=pinat-ah-e=ri
y=atziri-te-paye]DE
3m.A=pay-REGR-IRR=3m.O 3m.poss=person-poss-PL
‘That’s why he doesn’t have money to pay his staff.’(That’s why that he has
money to pay his staff is not the case.)
4.113 Maatsi evankaripaye, kaari kimisantahirori.
tzimatsi evaNkari-paye kaari
kimisaNt-ah-i=ro=ri]MC
EXIST young.man-PL NEG.P pay.attention-REGR-REAL=3n.m.O=REL
‘There are young people, who don’t pay attention to it.’ (lit. Those who pay
attention to it are not the case.’ )
(3) Intransitive negative polarity verb aritapaki ‘be enough’. The verb aritapaki has
been identified as such on the basis of its morphosyntactic properties. Its internal
structure is that of a verb, aritapaki ari-t-ap-ak-i [PP-EP-PRF-DIR-REAL], consisting of
the verb root ari ‘be the case’, the goal-oriented directional –ap which also expresses a
perfective sense, the perfective suffix –ak, and the realis status suffix –i. It functions as
head of intransitive clauses. The interpretation of the intransitive verb ‘be enough’ as a
negative polarity item reflects intuitions of language speakers who translate it as ‘no
more’, ‘basta.’ The verb has been attested with activity complement clauses in S
function. Examples are given in (4.114-4.116).
4.114 Aritapaki pikaataka.
[aritapaki]ME
[pi=kaa-t-ak-a]DE
be.enough
2S=bathe-EP-PRF-REAL
‘That you have bathed yourself in the river is enough.’
240
4.115 Aritapaki pimaki, tsame ahataite inaki.
[aritapaki]ME [pi=ma-ak-i]DE
tsame a=ha-t-a-it-e
ina-ki
be.enough
2S=sleep-PRF-REAL come.on 1PL.S=go-REGR-ICPL-IRR mother-LOC
‘That you have slept is enough, let’s go visit my mother.’
4.116
Aritapaki pitsoki noyaniri.
[aritapaki]DE [pi=tsok-i
no=kaniri]ME
be.enough
2A=pull.out
1SG.poss=manioc
‘That you have harvested my manioc is enough.’
Typically, the use of aritapaki as a complement-taking verb is meant to express
irritation and displeasure of the speaker who is fed up with the addressee’s behavior and
wants to stop the undesirable activity. In addition, aritapaki is used independently at the
end of narratives to signal the speaker’s intention to stop the narration.
(4) Evaluative verbs. Evaluative verbs kameetsataki kameetsa-t-ak-i [be.good-EP-PRFREAL] ‘be good’ and te oNkameetsate te o=N-kameetsa-t-e [NEG.REAL 3n.m.S=IRRbe.good-EP-IRR] ‘be not good ’are included here in the class of intransitive polarity
verbs. The syntactic function of both kameetsataki and te onkameetsate is that of heads of
intransitive clauses. The morphologically irregular positive polarity verb kameetsataki
‘be good’never marks subjects. The evaluation verbs generally occur before the
complement clause, as seen in (4.117-4.118), but may follow it, as seen in (4.119). In
(4.117), a potential complement clause is related to the negative polarity verb te
onkameetsate ‘be not good’ when the speaker refers to the inappropriateness of ridiculing
other beings such as plants and animals. Specifically, the speaker alludes to an
indigenous Amazonian conceptual framework of animism which extends human agency
to beings of other species including animals, plants, meteorological phenomena, some
artifacts, the dead, spirits, and gods. In (4.119), the activity complement clause precedes
241
the complement-taking verb. The context for this pragmatically-motivated clause
ordering is as follows. After the healer finished her description of a steam-bathing
technique, she concluded with a rhetorical question as if she were asking her audience for
their opinion. A few moments later, in her response to the question, she used a positive
evaluative verb with the fact complement clause in (4.118), evidently pleased with the
outcome of her demonstration.
4.117 Pairani te onkameetsate akantavaishitya.
[pairani te
o=N-kameetsa-t-e]ME
before NEG.REAL 3n.m.S=IRR-be.good-EP-IRR
[a=kaNta-vai-ashi-t-ia]DE
1PL.S=do-EP-DUR-APPL.INT-EP-IRR
‘That we mock others wasn’t appropriate long ago.’
4.118 Kameetsataki nokantakiroka.
[kameetsa-t-ak-i]ME
[no=kaNt-ak-i=ro=ka]DE
be.good-EP-PRF-REAL 1SG.A=do-PRF-REAL=3n.m.O=ADV
‘That I did it here was good.’
4.119
Naari ironyaaka okaratzi nantzirika naari kameetsatakirika o te onkameetsate?
[naari ironyaaka o=kara-tz-i
n=aNt-tz-i=ri=ka
naari ]DE
I
now
3n.m.S=CAP.C-EP-REAL 1SG.S=do-EP-REAL=REL=Q I
[kameetsa-t-ak-i=ri=ka
o
te
o=N-kameetsa-t-e]ME
be.good-EP-PRF-REAL=REL=Q or NEG.REAL 3n.m.S=IRR-be.good-EP-IRR
‘Is it good or bad what I do?’(lit. ‘That what I do is good or bad?’)
(5) Modal verbs. Complement-taking modal verbs kaNt ‘be able’ and ontzimatye ‘be
necessary’ are attested only with potential complement clauses in O function, as seen in
(4.120-4.124). Both kaNt ‘be able’ and ontzimatye o=N-tzim-aty-e [3n.m.S=IRR-havePROG-IRR] ‘be necessary’ are morphologically defective since they generally occur with
the 3SG subject person marker. Ontzimatye has completely grammaticalized and has
242
become an invariant modal verb. Note that the modal verb kaNt takes the -i/-e class
inflection when it expresses the above values of participant-internal modality determined
by the actor’s internal condition (his/her cognitive, emotional, or physical ability).
4.120
Ari onkantake ayero nyaantsi arika anyaanate.
ari [o=N-kaNt-ak-e]ME
[ay-e=ro
nyaantsi]DE arika
PP 1PL.S=IRR-be.able-PRF-IRR 1PL.get-IRR=3n.m.O word
when
a=nyaana-t-e
1PL.S=study-EP-IRR
‘We will be able to learn the language if we study.’
4.121
Te onkante naaka nameniro.
[te
o=N-kaNt-e]ME
[naaka n=amen-e=ro]DE
NEG.REAL 3n.m.S=IRR-be.able-IRR I
1SG.A=take.care.of-IRR=3n.m.O
‘I wasn’t able to take care of her.’
4.122 Yonta ikantakiro tsonkiro ivani.
[yoNta i=kaNt-ak-i=ro]ME
[tsoNk-i=ro
i=vani]DE
that
3m.A=be.able-PRF-REAL=3n.m.O finish-REAL=3n.m.O 3m.poss=land
‘That (man) was able to finish [cleaning] his land.’
4.123 Eentsipaye ontzimatye yamitakoteri irora inkoyavaite.
[eentsi-paye o=N-tzim-aty-e]ME
[y=amitako-t-e=ri
irora
child-PL
3n.m.S=IRR-have-PROG-IRR 3m.A=help-EP-IRR=3m.O DEM
i=N-koya-vai-t-e
incha-shi-patsa-ini]DE
3m.S=IRR-gather-DUR-EP-IRR plant-CL:small.flat-small.part-DIM
‘It is necessary for the children to help gather leaves.’
4.124 Ontzimatye irora amontyanake.
[irora
a=moNty-an-ak-e]DE
[o=N-tzim-aty-e]ME
3n.m.S=IRR-have-PROG-IRR whachamacallit 1PL.S=cross.river-DIR-PRF-IRR
‘It is necessary for us to cross the river.’
(6) Phasal verbs. The phasal verbs iNt ‘begin’ and tsoNk ‘finish’, ‘end up’ take fact or
activity complement clauses in O function. The phasal verbs require co-referential
subjects in both clauses. Examples are provided in (4.125-4.126).
243
4.125 Otsonkakiro okasavaitanahi kooyaka.
[o=tsoNk-ak-i=ro]ME
[o=kasa-vai-t-an-ah-i
3n.m.A=finish-PRF-REAL=3n.m.O 3n.m.S=cook.manioc-DUR-EP-DIR-REGR-REAL
kooya=ka]DE
woman=DEM
‘The woman finished cooking manioc.’
4.126 Ari intanaka ikonyaatzi mamaro.
ari [iNt-an-ak-a]ME
[i=konyaa-tz-i
mamaro]DE
PP begin-DIR-PRF-REAL 3m.S=appear-EP-REAL owl
‘This is the case that owls began to appear [since then].’
In addition to phasal verbs, there are grammatical forms which modify the meaning of
the verbs by expressing the concept of inception. The verbal affixes –an ‘directional
source/temporal ingressive’34 and –ima(int) ‘ingressive aspect’ also serve to indicate the
entry into an action expressed by the verb. Examples are provided in (4.127-4.128).
4.127
Airo okinkivarotantanaka shintsini.
airo
o=kiNkivaro-t-aNt-an-ak-a
shintsini
NEG.IRR 3n.m.S=old-EP-APPL.REAS-DIR-PRF-REAL quickly
‘In order for her not to begin to age quickly.’
4.128 Imponyashitakya nomoyakaimaitaro.
iNponyashitakya no=moy-ak-imai-t-a=ro
afterwards
1SG.A=boil-CAUS-INCH-EP-REAL=3n.m.O
‘Then I will begin to boil it [the head of a coconut].’
34
See the classic Payne text (1982) for a discussion of Ashéninka directionals, which argues that meanings
of the directional goal –apa and directional source –an extend to a temporal sense, namely that the
temporal meaning of –an is non-finality whereas -apa expresses finality. In narratives, –an is used to
indicate the beginning of a time span, or to indicate the continuation or middle of the time span; it often
occurs in a series of verbs. The end of time span is indicated by –apa.
244
(7) Manipulative verbs shini ‘allow’, tsani ‘resist’, and amitako ‘help’. The verbs of
manipulation shini ‘allow’, tsani ‘resist’, ‘prevent’ and amitako ‘help’ take potential
complement clauses in O function, as seen in (4.129-4.131).
4.129 Pishinitero noncheke pisaimpikite.
[pi=shini-t-e=ro]ME
[no=N-chek-e
2A=allow-EP-IRR=3n.m.O 1SG.S=IRR-cut-IRR
‘Allow me to cut your firewood.’
4.130 Itsanyaro ante pankotsi.
[i=tsani-a=ro]ME
[aNt-e
3m.A=prevent-REAL=3n.m.O build-IRR
‘He resisted that we build a house.’
pi=saiNpiki-te]DE
2poss=firewood-poss
paNkotsi]DE
house
4.131 Tsame antzirite, amitakoterita avavisakotaheri.
[tsame a=N-tziri-t-e
amitako-t-e=ri=ta]ME
come.on 1PL.S=IRR-be.person-EP-IRR help-EP-IRR=3m.O=OPT
[a=v-avis-ako-t-ah-e=ri]DE
1PL.A=CAUS-pass-APPL-EP-REGR-IRR=3m.O
‘Come on, we’ll turn into people, we’ll help liberate him [the Inka].’
A causer participant can be added without a complement clause filling an argument
slot. Specifically, verbal affixes with manipulative semantics such as causal prefixes
‘malefactive causative’ mi(n)-, ‘agentive causative’ o-~oi-~ov-, and sociative causative
suffix –ak serve to modify the meaning of the verb (for discussion and examples of
causative verbal affixes see §3.4.2).
4.3 Adverbial clauses
In the preceding section §4.2, I have discussed complement clauses which function as
core verb arguments. In this section, I will describe adverbial clauses functioning as verb
adjuncts which modify entire clauses. In this presentation, I largely draw on the semantic
245
types of clause linking proposed by Dixon (2010b: 2). Specifically, this analysis will
focus on the grammatical ways which Ashéninka Perené utilizes to link adverbial clauses
based on three main semantic types of adverbial clause linking: temporal (succession,
simultaneity, anteriority), conditional (possible and counterfactual), and consequence
(cause, purpose, result, and possible consequence). In syntactic terms, this account
makes a distinction between the main clause which can stand on its own and the
subordinate clause, adjoined to the main clause. In Ashéninka Perené, clause ordering is
not fixed: each semantic type of clause linking exhibits specific ordering patterns. In
§4.3.1, temporal adverbial clauses are outlined; §4.3.2 describes possible and
counterfactual conditional adverbial clauses35; §4.3.3 is an account of adverbial clauses
formed via consequence clause linking devices. Other clause linking types (contrast,
disjunction, and manner) are addressed in §4.3.4.
4.3.1 Temporal clause linking
There are three basic types of temporal linking: (i) temporal succession, (ii) temporal
simultaneity, and (iii) temporal anteriority. Table 41 is a summary of temporal clause
linking devices.
(i)Temporal succession.Temporal succession is expressed either by juxtaposition or by
the sequencing connectives ironyaaka ‘now’, ‘then’ and oponya/iponya ‘afterwards’.
When two clauses occur together in a sentence, the clause ordering often signals that the
event described in the first clause occurs before the event described in the second clause.
An illustration of the iconic order of the temporally sequenced clauses is given in (4.132).
35
Note that temporally sequenced and counterfactual conditional clauses are coordinated in Ashéninka
Perené.
246
TABLE 41. Temporal clause linking devices
Type of
temporal clause
linking
Succession
Form
Gloss
Function
Other functions
i/ponya
Overlap
arika
‘afterwards’,
‘then’
‘when’
=ra
‘when’
expresses sequence of
events
expresses a short
temporal overlap
expresses a short
temporal overlap
ovakera
‘when’
i/okaNta
‘in the
meantime’
irohatzi
‘until’
expresses temporal
anteriority
tekira
‘before’
expresses temporal
anteriority
lexical verb pony
‘come from’
conditional
connective ‘if’
non-subordinating
spatial (locativedemonstrative)
clitic ‘there’
non-subordinating
adverb ‘recently’
anteriority ‘until’ ;
succession ‘after
some time’
overlap ‘when’;
anteriority ‘before’;
spatial direction ‘up
to’;
negative polarity
adverb ‘not yet’
Anteriority
expresses a prolonged
temporal overlap
expresses a prolonged
temporal overlap
4.132 Naakiro kotsiro, nantaitaki inchatokira.
[n=a-ak-i=ro
kotsiro]MC [n=aNtai-t-ak-i
1SG.A=choose-PRF-REAL=3n.m.O machete 1SG.S=go.up-EP-PRF-REAL
inchato-ki=ra]MC
tree-LOC=DEM
‘I picked up the machete, and then climbed up the tree.’
As mentioned above, clausal linkers ironyaaka ‘now’, ‘then’ and iponya/oponya
‘afterwards’ signal temporal succession of events. The sequencing connectives occur in
the second conjunct. The semantics of the temporal adverb ironyaaka which generally
indicates that the action is taking place at the present moment extends to a sequencing
247
sense36. When ironyaaka occurs in narratives which describe sequenced past events, its
function is to signal the new turn of events or to introduce a new topic. In this role,
ironyaaka essentially performs a role of the sequencing connective, as seen in (4.133).
4.133 Oshiyakantzi oime pokahi, ironyaaka yantaitashitakiro henoki.
[o=shiyakaNt-tz-i
pok-ah-i]MC
3n.m.S=think-EP-REAL come-REGR-REAL
[ironyaaka y=aNtai-t-ashi-t-ak-i=ro
henoki]MC
then
3m.A=go.up-EP-APPL.INT-EP-PRF-REAL=3n.m.O top
‘She thought that her husband had come back, then he climbed up to where she
was.’
Additionally, the language has a group of sequencing connectives derived from the
verb pony ‘come from’ (see §3.4.4 for the examples of sequencing adverbs)37. They
minimally take the realis suffix –a and the subjective 3 person clitics i= /o= depending
on the gender of the actor participant in the preceding clause. Other derivational and
inflectional verb morphology include the directional suffix –an, perfective -ak, the
applicative of intention –ashi, and the irrealis morpheme N-..-ia. Data suggest that the
sequencing pony-forms are becoming morphologically irregular since subject proclitics
are often elided in everyday speech. It should be noted that when some language
consultants rechecked the accuracy of the transcriptions of the collected Ashéninka
Perené texts, they rejected the omission of subject proclitics as ‘incorrect’ on the grounds
36
Anderson observes that the semantics of ironyaaka extends beyond a ‘now’- temporal reference since it
often occurs in narratives about past events; specifically, when ironyaaka is found with intransitive verbs
whose subjects are marked as objects, it serves a pragmatic function of “focusing the main points of the
conversation” (1991:71-2).
37
A useful discussion of two Ashéninka adverbial vebs i/oponya and i/okanta is in Anderson (1983:79-90).
The Anderson account of the adverbial connectives states that i/oponya is a sequencing connective carrying
an aspectual sense of perfectivity while i/okanta is a contemporaneous marker indicating simultaneity of
two events; i/okanta generally indicates stativity of the event or state.
248
that this innovative phenomenon is ‘the result of the language contact with other
Ashéninka varieties’. However, the regularity of the subject proclitic elision in the
collected corpus shows that this phenomenon has become a common speech habit among
Ashéninka Perené speakers. The sequencing connective iponya/oponya may refer to a
point of time in the past, as shown in (4.134), or in the future, as seen in (4.135-4.136).
4.134 Naaki pashini, ponyashitaka aisatzi nochekakiro.
[n=a-ak-i
pashini]MC ponyashitaka aisatzi
1SG.S=pick-PRF-REAL other
then
also
no=chek-ak-i=ro]MC
1SG.A=cut-PRF-REAL=3n.m.O
‘I chose another branch, then I also cut it.’
4.135 Yoisotakiri shivitsaki, ponyashitaka yaaki shitashintsi aisatzi yamponatakiri.
[y=o-kiso-t-ak-e=ri
shivitsa-ki]MC
[ponyashitaka
3m.A=CAUS-be.hard-EP-PRF-IRR=3m.O vine.species-LOC then
y=a-ak-e
shitashintsi aisatzi y=aNponat-ak-e=ri]MC
3m.S=take-PRF-REAL mat
also
3m.A=wrap-PRF-IRR=3m.O
‘They will tie up the corpse with a vine, afterwards they will take a mat and also
wrap it around the corpse.’
4.136 Oponya nonkirikakero ampee, oponya naaka nontsakakero,
[oponya no=N-kirik-e=ro
aNpee]MC [oponya naaka
then
1SG.A=IRR-weave-IRR=3n.m.O cotton
then
I
no=N-tsak-ak-e=ro]MC
1SG.A=IRR-dye-PRF-IRR=3n.m.O
‘Afetrwards I will spin cotton, then I will dye it.’
(ii) Temporal simultaneity. Temporal simultaneity refers to an overlap of a time event in
one clause with the event’s time span in another clause. Simultaneity is expressed via the
temporal conjunction arika ‘when’, temporal clitic =ra ‘when’, temporal adverb ovakera
with the basic meaning ‘recently’, or the adverbial contemporaneous verb okaNta/ikaNta
249
‘in the meantime’. To indicate a brief temporal overlap of the non-past events, the
connective arika ‘when’ or the adverbial clitic =ra are employed; the former is used to
express a temporal overlap of non-past events, the latter is attested with overlapping past
events only. The temporal conjunction arika ‘when’ requires irrealis status in both main
and subordinate clauses to express a temporal overlap of the non-past events, as seen in
(4.137-4.138).
4.137
Arika ankivavetahyari, ero ivehaperonita.
[ero
[arika a=N-kiva-ve-t-ah-ia=ri ]SC
when 1PL.A=wash-FRUS-EP-REGR-IRR=3m.O NEG.IRR
i=veha-pero-ni-t-a]MC
3m.S=fade-AUG-AUG-EP-REAL
‘When we wash it, it won’t fade.’
4.138
Arika osantsatanake, nampinaikimaityaro.
[n=aNpinaik-imai-t-ia=ro]MC
[arika o=santsa-t-an-ak-e ]SC
when 3n.m.S=grow.big-EP-DIR-PRF-IRR 1SG.A=roll.up-INCH-EP-IRR=3n.m.O
‘When it becomes big, I roll it up.’
The arika-strategy is also used to denote a condition which can be met in the future.
Speakers comment that ‘when’ and ‘if’ interpretations of arika are in many circumstances
interchangeable, depending on the context of the speech situation. Formally, the
ambiguity can be resolved by the clause order. When the main clause precedes the
subordinate clause, it generally signals a lack of temporal connection between the two
events, suggesting the ‘if’-interpretation, as shown in (4.139).
4.139 Pivetsatapakyaro isha, arika pinyakero.
[pi=vitsa-t-ap-ak-ia=ro
isha]MC [arika pi=ny-ak-e=ro]SC
2A=greet-EP-DIR-PRF-IRR=3n.m.O grandma if
2A=see-PRF-IRR=3n.m.O
‘Say hi to grandma if you see her.’
250
Apart from the conjunction arika ‘when/if’, the temporal clitic =ra ‘when’ is
employed to signal a brief overlap in the events or states described in the main and
subordinate clauses. The ra-marked adverbial clauses are freely ordered, as exemplified
in (4.140-4.141).
4.140 Kamakiraha ikantzi, ‘Naaka te nokame.’
[kam-ak-i=ra=ha]SC
[i=kaNt-tz-i
naaka te
no=kam-e]MC
die-PRF-REAL=ADV=EMPH 3m.S=say-EP-REAL I
NEG.REAL 1SG.S=die-IRR
‘When he died, he said, ‘I didn’t die.’
4.141 Okantakanakina, isataitakinara aparoni aavintarontsi.
[o=kaNt-ak-an-ak-i=na]MC
3n.m.A=happen-CAUS-DIR-PRF-REAL=1SG.O
[i=sat-ai-t-ak-i=na=ra
aparoni aaviNtarontsi] SC
3m.A=stick-IMP.P-EP-PRF-EAL=1SG.O=ADV one
medicine
‘It did something to me when they administered a drug.’
In addition, the clitic =ra may exhibit a spatial meaning ‘where’ specifying the place
where the main clause event occurs, as seen in (4.142-4.143).
4.142 Osaikahetzira okovenkatzi; osheki manitzi.
[o=saik-a-he-tz-i=ra] SC
[o=koveNka-tz-i ]MC
osheki manitzi
3n.m.S=be.at-EP-PL-EP-REAL=ADV 3n.m.S=danger-EP-REAL many tiger
‘Where they lived, it was dangerous; there were many tigers [there].’
4.143
Ironyaaka antaroite manitzi ikenapaki onaryakara.
[ironyaaka antaro-ite manitzi i=ken-ap-ak-i]MC
CONN
big-AUG tiger
3m.S=walk-DIR-PRF-REAL
[o=nary-ak-a=ra]SC
3n.m.S=lie.prostrate-PRF-REAL=ADV
‘A huge tiger walked into the area where the women were lying [sleeping].’
A more prolonged overlap of activities or states in both clauses can be marked by the
temporal adverb ovakera ‘recently’ or the adverbial gender-sensitive contemporaneous
251
verb okaNta/ikaNta ‘in the meantime.’ The temporal adverb ovakera ‘recently’, when
joined to the subordinate clause, functions as the conjunction ‘when’. The connective
ovakera ‘when’ occurs only with past events. It is always preposed to the subordinate
clause predicate, as exemplified in (4.144-4.145).
4.144 Ovakera tsopachani aiki, ovakera eentsitapaki, airo ookavaishitaro.
[ovakera tsop-acha=ni
aiki ovakera eentsi-t-ap-ak-i]SC
when
pull.out-STAT=REL tooth when
child-EP-DIR-PRF-REAL
[airo
ooka-vai-ashi-t-a=ro]MC
NEG.IRR
throw.away-DUR-APPL.INT-EP-REAL=3n.m.O
‘When the first tooth is ready to fall out when we are little, we shouldn’t throw it
away thoughtlessly.’
4.145
Ari okimita pairani aisatzi naari, ovakera chooki oshitovapahi iraani.
[ari o=kimi-t-a
pairani aisatzi naari ]MC [ovakera chooki
PP 3n.m.S=be.similar-EP-REAL before also
I
when
sister
o=shitov-ap-ah-i
iraani]SC
3n.m.S=leave-DIR-REGR-REAL blood
‘The same happened to me long ago, when my sister menstruated for the first
time.’
The adverbial verb ikaNta/okaNta ‘while’, ‘in the meantime’ is derived from the
polyfunctional verb kaNt ‘be’, ‘happen’, ‘be able’, ‘say’. Morphological possibilities of
the adverbial verb kaNt are generally limited to the subject 3person proclitics i=/o= and
the realis marker -a. The gender-sensitive adverbial verb shows regular grammatical
agreement with the actor participant carrying out an action or undergoing a state in the
subordinate clause. It normally precedes the subordinate clause verb, as seen in (4.1464.147).
252
4.146
Ikanta ishinkitakeri, ironyaaka ashoshira kiyaki iroori.
[ikaNta i=shiNki-t-ak-i=ri]SC
[ironyaaka ashoshi=ra
CONN 3m.A=get.drunk-EP-PRF-REAL=3m.O now
armadillo=DEM
kiy-ak-i
iroori]SC
dig-PRF-REAL she
‘While the men were getting him drunk, the armadillo woman dug a hole.’
4.147 Okanta ironyaaka yora ashoshira ovitsikaki, ikantakiro iritsiro.’
[okaNta ironyaaka yora ashoshi=ra
o=vitsik-ak-i]SC
CONN now
DEM armadillo=DEM 3n.m.S=make-PRF-REAL
[i=kaNt-ak-i=ro
iritsiro]MC
3m.A=say-PRF-REAL=3n.m.O his.sister
‘While the female armadillo dug the hole, the [male] armadillo told his sister.’
The adverbial verb ikaNta/okaNta typically occurs in narratives about past events.
When repeated, it signals a lack of immediate succession of events when a prolonged
time period passes before the activity or state in the second clause eventuates. In this
function, ikaNta/okaNta can be translated as ‘after some time passed’, ‘until’ as seen in
(4.148-4.149).
4.148 Intaani yayeniri tzina, okanta, okanta amatavitakeri maini, kamaki.
[iNtaani y=ay-i=ni=ri
tzina]MC [okaNta okaNta
solely
3m.A=get-REAL=3O=3m.O pifayo
CONN CONN
amatavi-t-ak-i=ri
maini kam-ak-i ]SC
deceive-EP-PRF-REAL=3m.O bear die-PRF-REAL
‘He only brought her the pulp of pifayo (palm.sp.) to eat, after some time passed,
she deceived the bear [by pretending that] she was dead.’
4.149 Ishetaka, ishetaka, ikanta, ikanta kamaki.
[i=shet-ak-a
i=shet-ak-a] MC
[ikaNta ikaNta kam-ak-i]MC
3m.S=clean-PRF-REAL 3m.S=clean-PRF-REAL CONN CONN
die-PRF-REAL
‘He cleaned and cleaned his face, then he died.’
253
(iii) Temporal anteriority. The language does not have dedicated conjunctions or
formatives to express clause linking of temporal posteriority, which would correspond to
English conjunctions ‘since’ and ‘after’. However, temporal posteriorty can be indicated
by the iconic clause ordering, by the sequencing connective iponya/oponya ‘afterwards’
or by the temporal adverb pairani ‘long ago’ which can be employed to signal the endpoint of a posterior event, as seen in (4.150).
4.150 Makako irotaki notzimakotaki pairani.
no=tzim-ako-t-ak-i
pairani]SC
[makako]MC [irotaki
macaque
FOC
1SG.S=be.born-APPL-EP-PRF-REAL long.ago
‘Macaque this [name has been] since I was born.’
Temporal anteriority of events is coded with the syntactically independent elements
irohatzi ‘until’ and tekira ‘before’. The temporal anteriority linker irohatzi ‘until’(some
speakers pronounce it as irosatzi) marks an endpoint of the event in the subordinate
clause, projected into the future with respect to the event carried out in the main clause.
The clausal linker irohatzi ‘until’ can take the emphatic clitic =ta which attaches
indiscriminately to various hosts. When the clitic =ta is joined to irohatzi ‘until’, it
acquires the intensifying meaning ‘until the very last moment’. When the clausal linker
irohatzi ‘until’ is employed, the subordinate clause verb is obligatorily marked by the
applicative suffix of reason/instrumentality -aNt. Evidently, the applicative –aNt, which
is realized on the subordinate clause verb, signals the resultative aspect of the action
carried out in the subordinate clause38. The subordinate clause verb, when preceded by
irohatzi ‘until’, may appear in irrealis, as seen in (4.151-4.152), or in realis, as seen in
38
Payne observes that irohatzi always co-occurs with the applicative of reason –aNt and the relative marker
=ri (1989:286-7). In Ashéninka Perené, the collected corpus provides only evidence to support the
mandated co-occurrence of irohatzi with the applicative suffix –aNt marked on the subordinate clause verb.
254
(4.153-4.154), depending on the temporal frame of the main clause verb. Generally, a
realis event in the main clause requires the corresponding realis subordinate clause, and
vice versa. The main clause is generally followed by the irohatzi-initial subordinate
clause.
4.151 Ompasatatyero tsinki, tsinki, tsinki irohatzi ontsonkantakyaro.
[o=N-pas-a-t-aty-e=ro
tsiNki tsiNki tsiNki]MC [irohatzi
3n.m.S=crush-REP-EP-PROG-IRR=3n.m.O IDEO IDEO IDEO
until
o=N-tsoNk-aNt-ak-ia=ro]SC
3n.m.A=IRR-finish-APPL.REAS-PRF-IRR=3n.m.O
‘She will have to crush cotton until she finishes it [the pile].’
4.152 Onkantero irohatzita oshitovantahya.
[o=N-kaNt-e=ro]MC
[irohatzi=ta o=shitov-aNt-ah-ia]SC
3n.m.A=IRR-do-IRR=3n.m.O until=EMPH 3n.mS=leave-APPL.REAS-REGR-IRR
‘She will weave it until she leaves [her secluded place].’
4.153 Okaratzi mava ivaitakina iraantsive irohatzi noshitakotantaha.
[o=kara-tz-i
mava i=v-ai-t-ak-i=na
3n.m.S=COP.CAP-EP-REAL three 3m.A=put-IMP.P-EP-PRF-REAL=1SG.O
iraantsi=ve]MC [irohatzi no=shitako-t-aNt-ah-a]SC
blood=EXCL
until
1SG.S=get.well-EP-APPL.REAS-REGR-REAL
‘They administered three doses of blood until I got better.’
4.154
Ari isaikahetzi ikenkitsavaitaiyeni irohatzi ahetantapakari ivochokini.
[ari i=saik-a-he-tz-i
i=keNkitsa-vai-t-aiy-i-ni]MC
[irohatzi
PP 3m.S=be.at-EP-PL-EP-REAL 3m.S=talk-DUR-EP-PL-REAL-PL until
aree-t-aNt-ap-ak-a=ri
i=pochoki-ni]SC
arrive-EP-APPL.REAS-DIR-PRF-REAL=REL 3m.poss=dream-poss
‘They kept talking until both succumbed to sleep.’
The clause linker irohatzi exhibits other temporal extensions such as ‘before’ or
‘when’, as seen in (4.155-4.156). In addition, as a non-subordinating device, it occurs
255
with other temporal adverbs, as given in (4.157), or with nouns when it acquires a spatial
(directional) sense, as shown in (4.158).
4.155 Tsiirishiki ikitsataitakina kitamarori, irohatzi nohatantakya.
[Tsiirishi-ki
i=kitsa-t-ai-t-ak-i=na
kitamarori]MC
La Merced-LOC 3m.A=dress-EP-IMP.P-EP-PRF-REAL=1SG.O white
[irohatzi no=ha-t-aNt-ak-ia]SC
until
1SG.S=go-EP-APPL.REAS-PRF-IRR
‘They dressed me in a white gown in La Merced, before I went off [to Oroya].’
4.156
Ari ikantakotari mamaro, ikoshitzi aparoni kooyaniki irohatzi yatzirivetani.
[ari i=kaNt-ako-t-a=ri
mamaro i=koshi-tz-i
PP 3m.S=say-APPL-EP-REAL=3m.O owl
3m.S=kidnap-EP-REAL
aparoni kooya-aniki]MC [irohatzi y=atziri-ve-t-a-ni]SC
one
woman-DIM
until
3m.S=person-FRUS-EP-REAL-AUG
‘This is what they say about the owl which kidnapped a young woman when he
was a person long ago.’
4.157
Irotaki ikamantahetziri ireentsitepaye irohatzita ironyaaka.
irotaki i=kamaNt-a-he-tz-i=ri
ir=eentsi-te-paye
FOC 3m.A=advise-EP-PL-EP-REAL=3m.O 3m.poss=child-poss-PL
irohatzi=ta
ironyaaka
until=EMPH
now
‘This is what they have told their children until today.’
4.158 Ironyaaka shiramparika yaanakiro eentsi irohatzi opatsishiki shina.
ironyaaka shiraNpari=ka y=a-an-ak-i=ro
eentsi
now
man=DEM
3m.A=take-DIR-PRF-REAL=3n.m.O child
irohatzi o=patsishi-ki
shina
until
3n.m=top.branch-LOC kapok.tree
‘Then this man took the girl up to the top branch of the kapok tree (Ceiba
Pentandra).’
The clause-linking device tekira ‘before’ is used to indicate the endpoint of some
future activity with regard to the main clause event. The adverbial connective tekira has
the basic meaning of imperfectivity ‘not yet’, but when linked with the following clause,
256
tekira generally indicates ‘when X has not yet done Y’ or ‘before X does Y’. The
conjunction is composed of the negative polarity realis particle te, locative marker –ki,
and the adverbial marker =ra. The optative modal clitic =ta which expresses requests,
commands, possibility, a wish for something to come about, or an action that has not yet
occurred, is frequently realized on the clausal linker tekira to denote a yet unrealized,
hypothetical action subsequent to an end-point of another action carried out in the main
clause. Note that the subordinate clause verb is obligatorily marked by both the irrealis
status marker and the modal clitic =ta to express the hypothetical possibility of an action.
Examples are provided in (4.159-4.161).
4.159 Tekira inteterita, inkaatavaherita kaminkari ihataheta kametsa.
[tekira i=N-tet-e=ri=ta]SC
not.yet 3m.A=IRR-fill-IRR=3m.O=OPT
[i=N-kaa-t-av-ah-e=ri=ta
kamiNkari
3m.A=IRR-bathe-EP-DIR-REGR-IRR=3m.O=OPT dead
i=ha-t-ah-e=ta
kametsa]MC
3m.S=go-EP-REGR-IRR=OPT well
‘Before placing him [the corpse] in the box, they wash the dead person so that he
goes well.’
4.160 Apaitanaki eentsitatyai, tekirata aapitanaketa.
[a=pait-an-ak-i
eentsi-t-aty=ai]MC
[tekira=ta
1PL.S=get.white.hair-DIR-PRF-REAL child-EP-PROG=1PL.O not.yet=OPT
aapi-t-an-ak-e=ta]SC
elder-EP-DIR-PRF-IRR=OPT
‘We get white hair being young before we get old.’
257
4.161 Tekira pariitzimotyarita, ari irorave, pihatake, ontzimatye
pintyerovashitapakyari.
[ari
[tekira p=arii-tz-imo-t-ia=ri=ta]SC
not.yet 2A=arrive-EP-APPL.PRES-EP-IRR=3m.O=OPT PP
irora=ve
pi=ha-t-ak-e
o=N-tzim-aty-e
whachamacallit=EXCL 2S=go-EP-PRF-IRR 3n.m.S=IRR-have-PROG-IRR
pi=N-tyerov-ashi-t-ap-ak-ia=ri]MC
2S=IRR-kneel.down-APPL.INT-EP-DIR-PRF-IRR=3m.O
‘Before you arrive in his presence, you will go and you will have to kneel down.’
4.3.2 Conditional clause linking
Two varieties of conditional clause linking are distinguished: (i) possible conditional,
which involves the subordinate condition clause, and the main clause which describes a
possible result, e.g. ‘If I go hunting, I won’t eat anything’; and (ii) counterfactual
conditional which refers to a condition which might have been met in the past but for
some reason wasn’t; it involves the subordinate condition clause and the main clause
which describes the event that didn’t take place, e.g. ‘Had Beetle obeyed his brother-inlaw, his brother-in-law would have helped him.’ Table 42 below is an overview of
conditional linking devices.
(i) Possible conditional linking.The grammatical means for coding possible conditional
linking include the connector aririka, derived from the positive polarity verb ari ‘be the
case’ and the conditional clitic =rika. Like other adverbial conjunctions, aririka is
positioned clause-initially. This type of conditional linking always requires irrealis status
in both main and subordinate clauses, as seen in (4.162-4.163).
258
TABLE 42. Possible and counterfactual clause linking devices
Type of
condition
Possible
Form
Function
Other functions
aririka
expresses possible condition
‘if’
expresses possible condition
‘if’
expresses possible condition
‘if’
temporal simultaneity ‘when’
arika
=rika
terika
airorika
Counterfactual
=mi
negates a realis condition
clause ‘if not’
negates an irrealis condition
clause ‘if not’
marks counterfactual
condition
temporal simultaneity ‘when’
interrogative complement clause
‘if’;
comparison ‘like’
marker of clausal disjunction ‘or’
marker of clausal disjunction ‘or’
marker of an averted event;
marker of frustrative modality;
affective marker in requests
4.162
Aririka pimishakena, kitamaro novatsa.
[aririka pi=mish-ak-e=na]SC
[kitamaro no=vatsa]SC
if
2A=remove.skin-PRF-IRR=1SG.O white 1SG.poss=meat
‘If you remove my skin, my flesh is white.’
4.163
Aririka amonkaratakero maaroni kamenaantsika, aririka omonkaratapakya
atsokantyarori, ari anyaakero osheki ovatsa kaniri.
[aririka a=moNkara-t-ak-e=ro
maaroni kamenaantsi=ka
if
1PL.A=measure-EP-PRF-IRR=3n.m.O all
advice=DEM
aririka a=moNkara-t-ap-ak-ia
if
1PL.A=measure-EP-DIR-PRF-IRR
a=tsok-aNt-ia=ro=ri]SC
[ari a=ny-ak-e=ro
1PL.A=pull.out-APPL.REAS-IRR=3n.m.O=REL PP 1PL.A=see-PRF-IRR=3n.m.O
osheki o=vatsa
kaniri]MC
many 3poss=mass manioc
‘If we follow this entire advice, if we follow it when we harvest manioc, we’ll see
a big quantity of manioc.’
As mentioned above in my discussion of arika ‘when/if’, there is a close association
between the ‘when’ and ‘if’ clause linkings. In the same vein, the aririka-strategy is used
for coding the linking of temporal simultaneity. In (4.164-4.165), there is a clear temporal
259
connection between the main and subordinate clauses, which makes a ‘when’ translation
of aririka a preferred choice.
4.164
Aririka pitzimakakero eentsi aisatzi ontzimatye pintzinae kapichikitaite.
[aririka pi=tzim-ak-ak-e=ro
eentsi]SC [aisatzi oNtzimatye
when 2S=have-CAUS-PRF-IRR=3n.m.O child
also
be.necessary
pi=N-tzin-a-e
kapichi-kitaite]MC
2S=IRR-get.up-REGR-IRR small-be.morning
‘When you have your wife bear children, you will also have to get up early.’
4.165 Pantakotapakeri, aririka ishitovanaki kanaiki, ari pivanakeri oponkitziki
inchatoka.
p=aNt-ako-t-ap-ak-e=ri
[aririka i=shitov-an-ak-e
2A=hit-APPL-EP-DIR-PRF-IRR=3m.O when
3m.S=escape-DIR-PRF-IRR
kanai-ki]SC
tree.sp-CL:social.insect
[ari
PP
pi=v-an-ak-e=ri
2A=place-DIR-PRF-IRR=3m.O
o=poNkitzi-ki
inchato=ka]MC
3n.m.poss=ground-LOC tree=DEM
‘Bang on the tree, when ants come out, leave the bird on the ground by the trunk.’
In additon to the syntactic marker aririka, the conditional clitic =rika is used to code
possible condition clause linking. The clitic is attested with various hosts including
lexical verbs, as seen in (4.166), copula verbs, as seen in (4.167), and focused verbalized
pronouns, as seen in (4.168). The conditional clitic =rika can co-occur with syntactic
markers of condition/temporal succession aririka ‘when/if’, as seen in (4.166), or arika
‘when/if’, as shown in (4.168). Additionally, the conditional clitic =rika functions as the
marker of an interrogative complement clause ‘if’, as seen in (4.168).
260
4.166 Aririka ahate ashimatya nihaaki, anyapakerika ishiyanaki paampatzi, ari ayoteri
tzimatsira shima.
[aririka
a=ha-t-e
a=shima-t-ia
nihaa-ki
if
1PL.S=go-EP-IRR 1PL.S=fish-EP=IRR river-LOC
a=ny-ap-ak-e=rika
i=shiy-an-ak-e
1PL.S=see-DIR-PRF-IRR=COND 3m.S=run-DIR-PRF-IRR
paaNpatzi]SC
bird.species
[ari a=yo-t-e=ri
tzimatsi=ra
shima]MC
PP 1PL.S=know-EP-IRR=REL EXIST=ADV fish
‘If we go fishing to the river, if we see the bird paampatzi running around, it is the
case that we’ll know that there is fish there.’
4.167 Tzimatsirika pyarentsi, ahatzita ampakotavakeri.
[tzimatsi=rika pyarentsi]SC [ahatzi=ta
EXIST=COND manioc.beer also=EMPH
a=N-p-ako-t-av-ak-e=ri]MC
1PL.A=IRR-give-APPL-EP-DIR-PRF-IRR=3m.O
‘If there is manioc beer, we also should give it to him.’
4.168
Arika apavakiri, iritakirika sheripyari kovatsiri iyotai ankameetsatziririka, ari
itasonkakai.
[arika a=p-av-ak-e=ri
iritaki=rika sheripyari
if
1PL.A=give-DIR-PRF-IRR=3m.O FOC=COND shaman
kov-atsi=ri
i=yo-t=ai
want-STAT=REL 3m.A=know-EP=1PL.O
a=N-kameetsa-atziri-e=rika]SC
[ari i=tasoNk-ak=ai]MC
1PL.S=IRR-be.good-person-IRR=COND PP 3m.A=blow-PRF=1PL.O
‘If we give him [the food], and if he is a shaman who wants to know if we are
generous people, he will blow on us [he will bless us].’
The conditional marker =rika is also realized as a temporal marker ‘when’, as
exemplified in (4.169-4.170).
261
4.169 Iro oshitovahirika ari ipotsoitzirori ironyaaka ovoroki.
[iro o=shitov-ah-e=rika]SC
[ari
but 3n.m.S=leave-REGR-IRR=COND PP
i=potso-it-tz-i=ro=ri
ironyaaka o=poro-ki]MC
3m.A=paint-ICPL-EP-REAL=3n.m.O=REL now
3n.m.poss=face-LOC
‘But when she leaves, then they paint her [face].’
4.170 Hataki antamiki amine piratsi ivakotyarori oshitovahirika apaata.
[ha-t-ak-e
aNtami-ki
amin-e
piratsi
go-EP-PRF-REAL woods-LOC look.for-IRR animal
i=v-ako-t-ia=ro=ri]MC
[o=shitov-ah-e=rika
apaata]SC
3m.A=eat-APPL-EP-IRR=3n.m.O=REL 3n.m.S=leave-REGR-IRR=COND later
‘She will go to the woods to look for game that they will eat when she leaves her
seclusion.’
To negate a condition clause, the conditional negators airorika and terika are used.
The negators are composed of the negative particle, airo (irrealis) or te (realis), and the
conditional clitic =rika. The language makes a formal distinction between negated realis
and irrealis condition clauses. When the airo-negated condition clause refers to an
unrealized action or state, the subordinate clause verb occurs in realis mode whereas the
terika-negated clause, which signals that the action or state of the condition clause has
possibly taken place, requires the subordinate clause verb to be marked for irrealis. The
positive polarity main clause is marked for irrealis status to express a hypothetical
possibility of the expected result. The conditional negators generally precede the
condition clause verb, as seen in (4.171-4.173).
4.171
Airorika pinyiina nihaaki, ari nosaikeri kipatsiki.
[airorika
pi=ny-i=na
nihaa-ki]SC [ari
NEG.IRR.COND 2A=see-REAL=1SG.O water-LOC PP
no=saik-e=ri
kipatsi-ki]MC
1SG.S=be.at-IRR=REL ground-LOC
262
‘If you don’t see me in the water, I will be on the ground.’
4.172
Airorika ironyaaka nosaatantavaitzi, tekatsi nayeri noyari.
[airorika
ironyaaka no=saa-t-aNt-a-vai-tz-i]SC
NEG.IRR.COND now
1SG.S=bathe.in.hot.water-EP-CUST-DUR-REAL
[tekatsi
n=ay-e=ri
no=ya=ri]MC
NEG.EXIST 1SG.A=get-IRR=REL 1SG.S=eat.IRR=REL
‘If I don’t perform this steam-bathing procedure, there will be no way to get
food.’
4.173
Terika yaye kompiroshi, ari nopiyahi.
[terika
y=ay-e
koNpiroshi]SC [ari
NEG.COND 3m.S=get-IRR yarina
PP
‘In case he hasn’t cut yarina, I’ll come back.’
no=piy-ah-e]MC
1SG.S=return-REGR-IRR
The conditional negator terika also functions as a clausal operator of disjunction,
requiring the subjects of both the subordinate and main clauses to be co-referential, as
seen in (4.174-4.175).
4.174
Nokampitantantya vatsatsi terika namahe shima.
[no=kaNpitaNt-aNt-ia
vatsatsi]MC [terika n=am-ah-e
shima]MC
1SG.S=buy-APPL.REAS-IRR meat
or
1SG.S=bring-REGR-IRR fish
‘I’ll buy meat or will bring fish.’
4.175
Nohatatye Tsiirishiki terika nonkinanake Taaromaki.
[no=ha-t-aty-e
Tsiirishi-ki]MC [terika no=N-kin-an-ak-e
1SG.S=go-EP-PROG-IRR La Merced-LOC or
1SG.S=go-DIR-PRF-IRR
Taaroma-ki]MC
Tarma-LOC
‘I’ll be going to La Merced or may be going to Tarma.’
The clause which expresses an alternative often exhibits an elided verb, as exemplified in
(4.176).
263
4.176 Nopankitatye shiira terika katsimaropa.
[no=paNki-t-aty-e
shiira]MC
[terika katsimaropa]MC
1SG.S=plant-EP-PROG-IRR plantain.sp.
or
plantain.sp.
‘I’ll plant seeds of plantain seda or [I’ll plant] plantain isla.’
Terika can also express doubt or uncertainty when placed between two clauses,
approximating a meaning ‘maybe’. An example is provided in (4.177).
4.177 Noina ohatatzi tonkariki terika airo piyaha.
[no=ina
o=ha-t-atz-i
toNkari-ki]MC [terika airo
1SG.poss=wife 3n.m.S=go-EP-PROG-REAL hilltop-LOC maybe NEG.IRR
piy-ah-a]MC
return-REGR-REAL
‘My wife went to the hilltop [to work in the garden], maybe she will not be back
[today].’
(ii) Counterfactual conditional clause linking. Two types of counterfactual conditional
linking are distinguished. One type of counterfactual conditional linking is expressed by
the counterfactual clitic =mi which marks the verbs of both conjuncts, denoting a
hypothetical action or state which never took place and requiring irrealis inflection in
both clauses. Note that when either clause is negated, it takes realis inflection, as seen in
(4.178-4.179).
4.178 Yareetahyami kametsa, ero ivashiventaitzirimi ironyaaka Tziivito.
[y=aree-t-ah-ia=mi
kametsa]MC [ero
3m.S=arrive-EP-REGR-IRR=CNT.F well
NEG.IRR
i=pashiveNt-ai-tz-i=ri=mi
ironyaaka Tziivito]MC
3m.A=embarrass-IMP.P-EP-REAL=3m.O=CNT.F now
Beetle
‘Had they arrived safely [together], they wouldn’t have embarrassed Beetle.’
264
4.179 Yookaitanakiri irirori ipiyatsata, ero ipiyatsatami, ero yookaitzirimi.
y=ook-ai-t-an-ak-i=ri
irirori i=piyatsa-t-a
3m.A=abandon-IMP.P-EP-DIR-PRF-REAL=3m.O he 3m.S=disobey-EP-REAL
[ero
i=piyatsa-t-a=mi]MC
NEG.IRR 3m.S=disobey-EP-REAL=CNT.F
[ero
NEG.IRR
y=ook-ai-tz-i=mi]MC
3m.S=abandon-IMP.P-EP-REAL=CNT.F
‘They left Beetle because he had disobeyed; hadn’t he disobeyed, they wouldn’t
have left him.’
Another type of counterfactual clause linking includes an irrealis clause marked by
the counterfactual clitic =mi which has an avertive sense, denoting a closely missed
event. The second realis clause describes a circumstance which didn’t allow the
hypothetical action in the irrealis clause to eventuate. The marker of negated condition
airorika ‘if not’, which essentially functions as a marker of clausal disjunction, is used to
link such clauses, as seen in (4.180).
4.180 Apaatatziro isaiki, ironyaaka ipyakiri mapi airorika ironyaaka itsonkahetakaimi.
[apaatatziro i=saik-i
ironyaaka i=pi-ak-i=ri
later
3m.S=be.at-REAL now
3m.A=transform-PRF-REAL=3m.O
mapi]MC [airo=rika
ironyaaka i=tsoNka-he-t-aka-e=mi]MC
stone NEG.IRR=COND now
3m.S=finish-PL-EP-PRF-IRR=CNT.F
‘Later on he stayed there and transformed into a stone, or he would have killed
everyone.’
In single-clause statements, the counterfactual codition clitic =mi functions as a
marker of frustrative modality to express an action which failed to succeed or was
performed in vain. Generally, it signals that the speaker’s (or another participant’s)
expectations weren’t met. Examples are given in (4.181-4.182).
265
4.181
Intsompoikira ari nosaikimi kametsa.
intsoNpoi-ki=ra
ari
no=saik-i=mi
kametsa
inside-LOC=ADV there
1SG.S=be.at-IRR=CNT.F well
‘Inside [the corral] I was supposed to stay in comfort.’
4.182 Nonintzi nonkamemi.
no=niNt-tz-i
no=N-kam-e=mi
1SG.S=want-EP-REAL 1SG.S=IRR-die-IRR=CNT.F
‘I wanted to die [but it didn’t happen].’
In addition, the counterfactual marker =mi is used in affective requests. Examples are
provided in (4.183-4.184).
4.183 Ikantziro, ‘Isha, pimpinatyami naari.’
i=kaNt-tz-i=ro
isha
pi=N-pinat-ia=mi
naari
3m.A=say-EP-REAL=3n.m.O grandma 2S=IRR-give-IRR=CNT.F I
‘He said to her, ‘Grandma, please please give it to me.’
4.184
Ikantzi, ‘Tsametyami.’
i=kaNt-tz-i
tsame=tya=mi
3m.S=say-EP-REAL come.on=EMPH=CNT.F
‘He said, ‘Please let’s go.’
4.3.3
Consequence clause linking
Four subtypes of consequence linking have been attested: (i) causal linking which
involves the subordinate clause specifying the reason for the event or state carried out in
the main clause, e.g. ‘He plucked her eyes out because she had responded to his
greeting’; (ii) purposive linking which refers to the main clause describing the intentional
activity performed to ensure the accomplishment of another activity in the subordinate
clause, e.g. ‘They made her vomit in order to purge her of her laziness’; (iii) resultative
linking which marks the subordinate clause describing the result of an action carried out
in the main clause, e.g. ‘I gave him a steam bath, that’s why he is well now’; (iv) and
266
undesirable possible consequence linking which marks the subordinate clause as an
unwanted event whereas the main clause describes an action performed in order to avoid
this undesirable event, e.g. ‘You should get up early lest your wife hate you.’ Table 43 is
an overview of consequence clause linking devices.
TABLE 43. Consequence clause linking devices
Type of
consequence
clause linking
Causal
Purposive
Resultative
Undesirable
possible
consequence
Form
Gloss
Function
tema/kama
o/ikaNta
‘because’
‘because’
expresses cause
-aNt…=ri
‘in order to’
okaNta/oNkantya+
-aNt..(=ri);
okaNta/oNkantya
-ashi
‘so that X can/
could/might/should’
expresses a
deliberate
consequence
‘with the intent to’
‘in order to’
motion verbs
irotaki + -aNt..=ri
‘to’
‘that’s why’
=kari
‘lest’
Other
functions
simultaneiity
‘while’, ‘in the
meantime’
instrumental
sense ‘with X’
adverbial
marker ‘to
plan/intend to;
failed purpose
expresses an
unintended
consequence
expresses
unwelcome
possible
consequence
(apprehensive
marker)
marks a
focused clause
with an
instrumental
participant
single-clause
warning
marker
(i) Causal linking. Ashéninka Perené has grammaticalized few markers of causal linking
including the syntactic markers of causality tema or kama ‘because’ and the causal
connector okaNta/ikaNta with the meaning ‘because’. The semantically equivalent
267
markers of causality tema and kama ‘because’ commonly express causal linking in
Ashéninka Perené. As seen in (4.185-4.187), the position of the tema/kama-initial clause
is not restricted. In (4.185), a reference is made to the speaker’s belief that animals,
including deer, used to be humans; due to various circumstances, animals lost their
human form. This particular character, which pretended to be hunting deer in the woods,
was in fact cutting meat from his thighs and bringing it home to feed his family. When he
was outed as a liar, out of shame he transformed into a deer. In (4.186), the speaker’s
comment refers to an Ashéninka cultural conception of sorcery involving demonic
creatures like oovero ‘bird species’(called ‘the eye of the dead’) and human witches. It
was believed that the appearance of the singing oovero next to a house signaled that the
warriors were coming to kill its inhabitant who had been identified as a witch.
4.185 Te oshekite ivatsa ivoriki maniro temakya itotatziro ivatsa pairani.
te
osheki-t-e
i=vatsa
i=pori-ki
maniro
NEG.REAL be.much-EP-IRR 3m.poss=meat 3m.poss=thigh-LOC deer
tema=kya
i=tot-atz-i=ro
i=vatsa
pairani
because=EMPH 3m.A=cut-PROG-REAL=3n.m.O 3m.poss=flesh before
‘The thighs of the deer don’t have much meat because it had cut itself long ago.’
4.186 Pairani ironyaaka ishekitantakari oovero, kama ivayetziri ashaninkapaye
iitamatzitavakayeta.
[pairani ironyaaka i=sheki-t-aNt-ak-a=ri
oovero]MC
before now
3m.A=much-EP-APPL.REAS-PRF-REAL=REL hawk
[kama
ivayetziri a=shaniNka-paye
because warriors 1PL.poss=fellowman-PL
i=itamatzi-t-av-ak-aiy-e=ta]SC
3m.S=identify.as.a.witch-EP-DIR-PRF-PL-REAL=OPT
‘The numbers of oovero [birds which heralded death] increased in the old times
because warriors were killing my fellowmen who had been identified as
sorcerers.’
268
4.187 Tema kipatsi anatzi, kipatsi ampiyanahi.
[tema
kipatsi a=na-tz-i]SC
[kipatsi a=N-piy-an-ah-e]MC
because earth 1PL.S=be-REAL earth 1PL.S=IRR-return-DIR-REGR-IRR
‘Because we are of the earth, we’ll return to the earth.’
Additionally, causal linking is expressed by the polyfucntional verb kaNt ‘say’, ‘be’,
‘be able’, ‘do’, ‘happen’. As a causal connector, the verb takes a-inflection. It occupies
the subordinate clause-initial position and shows gender agreement with the subject of the
subordinate clause, as shown in (4.188-4.189).
4.188 Ikanta ironyaaka isamaitakari, ikantakiri Navireri.
ikaNta ironyaaka i=sam-ai-t-ak-a=ri]SC
[CONN now
3m.A=get.upset.with-IMP.P-EP-PRF-REAL=3m.O
i=kaNt-ak-i=ri
Navireri
3m.A=say-PRF-REAL=3m.O person’s.name
‘Because they were upset with him, they said to Navireri.’
4.189 Okanta otenkanakira nonintziro nompankite santari.
okaNta oteNkanaki=ra
no=niNt-tz-i=ro
CONN rocky.place=DEM 1SG.A=want-REAL=3n.m.O
no=N-paNki-t-e
saNtari
1SG.A=IRR-plant-EP-IRR cedar.tree
‘Because that is a rocky place, I want to plant cedar.’
Another common strategy is to indicate a causal relationship between two clauses by
leaving the clausal linking formally unexpressed. In (4.190), the first clause in the
biclausal string refers to the result, and the second clause refers to the cause.
4.190 Yookaitanakiri irirori, ipiyatsata.
[y=ook-ai-t-an-ak-i=ri
irirori]MC
3m.A=abandon-IMP.P-EP-DIR-PRF-REAL=3m.O he
[i=piyatsa-t-a]MC
3m.S=disobey-EP-REAL
269
‘They left him [Beetle] because he had disobeyed.’
(ii) Purposive linking.Three types of purposive linking are attested: (a) the purposive
construction with the applicative of reason/instrumentality –aNt;(b) the purposive
construction with the applicative of intent –ashi;and (c) the motion purpose construction.
I will discuss each type in detail below.
(a) The purposive construction with the applicative of reason/instrumentality –aNt. The
applicative of reason/instrumentality –aNt, co-occurring with the relativizer =ri, marks
the subordinate clause which describes the purposeful activity performed to accomplish
the action carried out in the main clause. Note that the applicative –aNt has a more basic
instrumental sense when it indexes the instrumental referent on the verb with the third
person clitics =ri (masculine) or =ro (non-masculine). Examples of the aNt-purposive
linking are given in (4.191-4.193).
4.191
Ponya ikaataitakiro ponkaro ovatsatantanakyari.
[ponya i=kaa-t-ai-t-ak-e=ro
poNkaro]MC
later 3m.A=bathe-EP-IMP.P-EP-PRF-IRR=3n.m.O bejuco
[o=vatsa-t-aNt-an-ak-ia=ri]SC
3n.m.S=get.fat-EP-APPL.REAS-DIR-PRF-IRR=REL
‘Afterwards she will be bathed with the herb of bejuco in order to get fat.’
4.192
Nantziro novolsatepaye antetantyari paitarikapaye pinteteri.
[n=aNt-tz-i=ro
no=volsa-te-paye]MC
1SG.A=make-EP-REAL=3n.m.O 1SG.poss=bag-poss-PL
[a=N-tet-aNt-ia=ri
paita=rika-paye pi=N-tet-e=ri ]SC
1PL.S=IRR-fill-APPL.REAS-IRR=REL WH=COND-PL 2S=IRR-fill-IRR=REL
‘I make my bags for us to fill them with anything that you will put in them.’
4.193
Nampitantaro onkameetsatantyari.
[n=aNpi-t-aNt-a=ro]MC
[o=N-kameetsa-t-aNt-ia=ri]SC
1SG.A=repeat-EP-INS-REAL=3n.m.O 3n.m.S=IRR=be.good-EP-APPL.REAS-IRR=REL
‘I repeatedly work with this to keep it straight.’
270
Positive polarity purposive clauses occur only in irrealis due to their volitional
semantics. However, when the purposive clause is negated, the purposive clause verb
takes realis inflection and the relativizer =ri is elided, as shown in (4.194).
4.194 Oshitovanahita pashini kametsa ovakerari oishi, airota okinkivarotzitantaha,
okitamarotzitantaha shintsini.
[o=shitov-an-ah-e=ta
pashini kametsa
3n.m.S=grow-DIR-REGR-IRR=OPT
other good
ovakerari o=ishi]MC
new
[airo=ta
o=kiNkivarotzi-t-aNt-ah-a
3n.m.poss=hair NEG.IRR=OPT 3n.m.S=be.old-EP-APPL.REAS-REGR-REAL
o=kitamarotzi-t-aNt-ah-a]SC
3n.m.S=be.white-EP-APPL.REAS-REGR-REAL
‘She will grow another, new, good hair, in order for her not to grow old, not to
grow grey.’
The positive and negative polarity aNt-purposive clauses often co-occur with the
morphologically irregular adverbial verb onkantya with the basic meaning ‘happen’. The
alternative form okaNta is used infrequently, based on the collected corpus. The adverbial
verb is derived from the verb kaNt ‘say’, ‘be’, ‘do’, ‘happen’, ‘be able’. Its inflectional
morphology is generally limited to the 3n.m.person marker o= and the discontinuous
irrealis status morpheme N- ..-ia (or the realis suffix –a). The verb, translated as ‘so
that’, intensifies the modal semantics of the purposive construction. When
okaNta/oNkaNtya is used, the verbal predicate expresses modal values of deontic or
epistemic modalities. In this function, the verb takes the -a/-ia realis status inflection and
precedes the purposive aNt-marked verb, as seen in (4.195-4.196). When the subordinate
clause verb is negated, the adverbial verb okaNta/oNkaNtya occupies the slot
immediately before the irrealis negative particle airo, as exemplified in (4.197).
271
4.195
Ironyaaka ariitapaha kitaiteri onkantya oshitovantahyari kooyaraha.
[ironyaaka arii-t-ap-ah-a
kitaiteri]MC [o=N-kaNt-ia
now
arrive-EP-DIR-REGR-REAL day
3n.m.S=IRR-happen-IRR
o=shitov-aNt-ah-ia=ri
kooya=ra]SC
3n.m.A=leave-APPL.REAS-REGR-IRR=REL woman=DEM
‘The day arrives so that the woman could leave her seclusion.’
4.196 Yapotohetakiri maaroni okanta yantavaitakantyariri.
[y=apoto-he-t-ak-i=ri
maaroni]MC [o=kaNt-a
3m.A=gather-PL-EP-PRF-REAL=3m.O all
3n.m.S=happen-REAL
y=aNtavai-t-ak-aNt-ia=ri=ri]SC
3m.A=work-EP-CAUS-APPL.REAS-IRR=3m.O=REL
‘He gathered them all so that he could make them work for him.’
4.197 Isataitavakina onkantya airo okatsikanta.
[i=sat-ai-t-av-ak-i=na]MC
[o=N-kaNt-ia
3m.A=stick-IMP.P-EP-DIR-PRF-REAL=1SG.O 3n.m.S=IRR-happen-IRR
airo
o=katsik-aNt-a]SC
NEG.IRR 3n.m.S=feel.pain-APPL.REAS-REAL
‘They gave me an injection so that my arm shouldn’t feel pain.’
(b) OkaNta/oNkaNtya-purposive linking. The clausal linker okaNta/oNkaNtya commonly
serves as a purposive linker on its own, apart from the purposive construction –aNt ..=ri.
In (4.198), it shows wider possibilities in its morphology by taking the emphatic clitic
=kya.
4.198 Ontzimatye ankaimakanteri iriri onkantyakya maaroni avitsikahero kameetsa.
[o=N-tzim-aty-e
a=N-kaim-ak-aNt-e=ri
3n.m.S=IRR-have-PROG-IRR 1PL.A=IRR-call-CAUS-APPL.REAS-IRR=3m.O
[o=N-kaNt-ia=kya
maaroni
ir=iri]MC
3m.poss=father 3n.m.S=IRR-happen-IRR=EMPH all
a=vitsik-ah-e=ro
kameetsa]SC
1PL.A=fix-REGR-IRR=3n.m.O well
‘We have to call his father on the phone so that we can resolve the issue
effectively.’
272
(c) The purposive construction with the applicative of intent –ashi.The semantic contents
of the purposive applicative –aNt and the applicative of intent –ashi are at first blush
similar in that that they both convey a general sense of purpose. However, there is a clear
distinction in the semantic and syntactic types of clauses they occur with; -aNt appears on
the verbal predicates of subordinate clauses which describe a focal purposeful activity,
e.g. [I repeat this]MC [in order to keep it straight]SC, whereas –ashi marks the main clause
verb which refers to the supporting activity, carried out to ensure the realization of the
focal purposeful activity described by the subordinate clause, e.g. [I repeat this with the
intent/in order]MC [to keep it straight]SC. As a result, the ordering of clauses marked by
either applicative is distinct. In general, the main clause with the ashi-marked verb
precedes the subordinate clause; the reverse clause ordering is typically attested with the
applicative –aNt when it occurs in the purposive function. Another formal characteristic
concerns realis/irrealis inflection of the verb marked by either purposive morpheme. As
mentioned above, the subordinate clause verb with the attached applicative –aNt takes the
irrealis suffixes –e or -ia. However, when the main clause verb occurs with the
applicative –ashi, it is generally marked by the realis status suffixes –i or -a39. Examples
of the ashi-purposive linking are given in (4.199-4.201).
39
The Payne (1989) analysis of the functions of the applicative of intent –ashi, based on the distinction
between reflexive and non-reflexive verbs, states that when –ashi has a purposive sense, it occurs only with
non-reflexive verbs, marked by the non-reflexive suffixes –i or –e; when –ashi appears with reflexive -a/ya-verbs, it has a meaning ‘by mistake/by chance/no more’ (267-8). The Ashéninka Perené corpus shows
that the functions of –ashi are not motivated by the reflexivity of the verb host that the morpheme attaches
to.
273
4.199 Ovashitatyaro ovaanyote airo okonyaatzitzi.
o=v-ashi-t-aty-a=ro
3n.m.A=put-APPL.INT-EP-PROG-REAL=3n.m.O
o=vaanyo-te
3n.m.poss=scarf-poss
airo
o=konyaa~tzitzi
NEG.IRR 3n.m.S=appear~REAL. EMPH
‘She wears a scarf [with the intent] not to show [her shaven head].’
4.200 Avashitakiro ashintsatakeri.
a=v-ashi-t-ak-i=ro
a=shintsa-t-ak-e=ri
1PL.S=place-APPL.INT-EP-PRF-REAL=3n.m.O 1PL.A=string-EP-PRF-IRR=3m.O
‘We combine with another thread [with the intent] to continue to string the beads.’
4.201 Naashitzimaitaro inchapataki nonkantero tyak, tyak.
n=a-ashi-tz-imai-t-a=ro
inchapataki
1SG.A=take-EP-APPL.INT-INCH-EP-REAL=3n.m.O spatula
no=N-kaNt-e=ro
tyak tyak
1SG.A=IRR-do-IRR=3n.m.O IDEO IDEO
‘I take a spatula [with the intent] to stir it [the water] chak, chak [sound of
splashing water].’
The applicative suffix –ashi can have other senses which are essentially revealed
contextually, on the basis of the pragmatics of a speech situation. For example, in a
single-clause statement, the applicative can have a sense ‘to intend to get X involved’. In
this function, the applicative suffix –ashi indexes the purposefully affected participant as
direct object on both transitive and intransitive verbs, as seen in (4.202).
4.202 Nokantashitari, ‘Oovero, paapaintero’.
no=kaNt-ashi-t-a=ri
oovero
1SG.A=say-APPL.INT-EP-REAL=3m.O bird.species
p=a-apaiNt-e=ro
2A=take-once-IRR=3n.m.O
‘I said [this] intending to jokingly ask him, ‘Oovero, take her.’
274
4.203 Okemakiri iraka, oshiyashitapakana.
o=kem-ak-i=ri
ir-ak-a
3n.m.A=hear-PRF-REAL=3m.O cry-PRF-REAL
o=shiy-ashi-t-ap-ak-a=na
3n.m.A=run-APPL.INT-EP-DIR-PRF-REAL=1SG.O
‘She heard him [the baby] cry and ran intending to help me.’
4.204 Ihatashitanakiri, inyapatziri isavoinataka iitsari.
i=hat-ashi-t-an-ak-i=ri
i=ny-ap-atz-i=ri
3m.A=go-APPL.INT-EP-DIR-PRF-REAL=3m.O
3m.A=see-DIR-PROG-REAL
i=savoina-t-ak-a
i=itsa-ri
3m.S=wear.a.cap.on.the head-EP-PRF-REAL 3m.poss=cushma-poss
‘They went intending to look for him and they found him wrapped in his
traditional robe.’
The appicative –ashi can indicate the intent or purpose which has not been achieved.
The failed action may be due to the actor’s lack of skill or unwillingness to cooperate, as
seen in (4.205-4.207).
4.205
Aririka amonkaratakero maaroni kamenaantsika, airo apiyashita aririka ahate
nihaaki aanakero ashimperi.
aririka a=moNkara-t-ak-e=ro
maaroni kamenaantsi=ka airo
when 1PL=measure-EP-PRF-IRR=3n.m.O entire advice=DEM
NEG.IRR
a=piy-ashi-t-a
aririka a=ha-t-e
nihaa-ki
1PL.S=return-APPL.INT-EP-REAL when 1PL.S=go-EP-IRR river-LOC
a-an-ak-e=ro
a=shiNperi
take-DIR-PRF-IRR=3n.m.O 1PL.poss=fish.trap
‘If we follow this advice, we won’t return empty-handed when we go to the river
and take our fish trap.’
4.206 Ironyaaka atsikashitaro oshetaki takik, takik.
ironyaaka atsik-ashi-t-a=ro
o=shetaki
takik takik
now
bite-APPL.INT-EP-REAL=3n.m.O 3n.m.S=fingernail IDEO IDEO
‘She pretended that she was biting her nails.’
275
4.207 Airo amenashitari ninkarika pokatsini inkivantai avankoki.
airo
amen-ashi-t-a=ri
niNkarika pok-atsi=ni
come-STAT=REL
NEG.IRR see-APPL.INT-EP-REAL=3m.O whoever
i=N-kivaNt=ai
a=paNko-ki
3m.A=IRR-visit=1PL.O 1PL.poss=house-LOC
‘We shouldn’t stare thoughtlessly at the person who comes to visit us in our
house.’
In (4.208-4.209), the suffix –ashi exhibits an adverbial sense ‘intentionally’ or ‘to
plan/intend to’ when occurring in the freely ordered biclausal strings containing both
purposive applicatives –aNt and –ashi.
4.208
Aka novetsikantayetarori, novashitayetziro aka.
[aka no=vetsik-aNt-a-ye-t-a=ro=ri]SC
here 1SG.A=make-APPL.REAS-EP-DIST-EP-REAL=3n.m.O=REL
[no=v-ashi-t-a-ye-tzi=ro
oka]MC
1SG.A=place-APPL.INT-EP-EP-DIST-REAL=3n.m.O DEM
‘In order for me to make it, I plan to combine it [the tree bark with the fabric].’
4.209
Oponya naashitakero patsitaroki irora nopotsotantakyarori.
[oponya n=a-ashi-t-ak-e=ro
patsitaroki]MC
then
1SG.A=take-APPL.INT-EP-PRF-IRR=3n.m.O tree.bark
[irora no=potsot-aNt-ak-ia=ro=ri]SC
DEM 1SG.A=dye-APPL.REAS-PRF-REAL=3n.m.O=REL
‘Then I will [plan to] take tree bark in order to dye the fabric.’
(d) The motion purpose construction. In this construction, the motion verbs ha ‘go’, shiy
‘run’, kin ‘walk’, pok ‘come’ and some others appear in the main clause describing some
motion carried out with the purpose of accomplishing a subordinate clause event. Since
the subordinate clause event is believed to be hypothetical, the subordinate clause verb is
marked as irrealis. The main clause rigidly precedes the subordinate clause. As shown in
(4.210-4.214), the subjects of the main and subordinate clauses are always coreferential.
276
4.210 Nohataki osheki nampitsi nonyaatsatya notsipatakari kisaakiripaye.
[no=ha-t-ak-i
osheki naNpitsi]MC [no=nyaatsa-t-ia
1SG.S=go-EP-PRF-REAL much place
1SG.S=play-EP-IRR
no=tsipa-t-ak-a=ri
kisaakiri-paye]SC
1SG.A=be.together-EP-PRF-REAL=3m.O black.person-PL
‘I traveled to other places to play football with blacks.’
4.211 Shiyanaka pashini notomi yamine kaaro.
[shiy-an-ak-a
pashini no=tomi]MC
run-DIR-PRF-REAL other 1SG.poss=son
‘My other son dashed to look for a car.’
4.212
[y=amin-e
kaaro]SC
3m.S=look.for-IRR car
Ikeni ivametairi onkantya yamitakoteri Juan Santos Atahualpa.
[i=vamet-a-e=ri
o=N-kaNt-ia
[i=ken-i] MC
3m.S=walk-REAL 3m.A=teach-REP-IRR=3m.O
3m.S=IRR-happen-IRR
y=amitako-t-e=ri
Juan Santos Atahualpa]SC
3m.A=help-EP-IRR=3m.O person’s.name
‘He came to teach them so that they could help Juan Santos Atahualpa.’
4.213 Ipoki ishimatya otapiki.
[i=pok-i]MC
[i=shima-t-ia
o=tapi-ki]SC
3m.S=come-REAL 3m.S=fish-EP-IRR 3m.poss=bottom-LOC
‘They come to fish to the bottom of the canyon.’
4.214 Ishonkanaharo otzishi yareetantahyari isaikahetzinta.
[i=shoNk-an-ah-a=ro
otzishi]MC
3m.A=go.around-DIR-REGR-IRR=3n.m.O hill
[y=aree-t-aNt-ah-ia=ri
i=saik-a-he-tz-i=Nta]SC
3m.S=arrive-EP-APPL.REAS-REGR-IRR=REL 3m.S=be.at-EP-PL-EP-REAL=ADV
‘They walked around the hill to arrive at where they lived.’
(iii) Resultative linking. By definition, the resultative subordinate clause refers to a
consequence of what is described in the main clause, e.g. [He brought us words [the
Bible]]MC, [that’s why we now can read and write]SC. The causal connection between the
main and subordinate clauses is somewhat tenuous since the main clause typically
involves a putative circumstance which could result in the outcome described by the
277
subordinate clause. The purposive and resultative linkings are marked by the same
morpheme, i.e. by the applicative of reason/instrumentality –aNt. When the applicative
appears in the resultative function, it has the meaning ‘as a result/consequence’, ‘that’s
why.’ The resultative clause verb is generally marked for realis by the suffixes –a and-i.
The resultative linking is performed in conjunction with the focused verbalized pronouns
irotaki/iritaki. In (4.215-4.217), positive and negative polarity resultative clauses are
formally equivalent except for the elided relativizer =ri on the negated subordinate verb,
as seen in (4.218).
4.215 Ipyanakeri ironyaaka kataripaye, irotaki ishekitantari katari aka.
[i=pi-an-ak-i=ri
ironyaaka katari-paye]MC [irotaki
3m.A=convert-DIR-PRF-REAL=3m.O now
wild.duck-PL FOC
i=sheki-t-aNt-a=ri
katari
aka]SC
3m.S=be.much-EP-APPL.REAS-REAL=REL wild.duck here
‘They converted in ducks, that’s why there are a lot of ducks here.’
4.216 ‘Tsame ampiiri tsamiri, osheki yantantzi,’ irotaki ipyantakariri tsamiri.
[tsame
a=N-pi-i=ri
tsamiri
osheki y=aNtaNt-tz-i]MC
come.on 1PL.A=convert-IRR=3m.O curassow much 3m.S=fight-EP-REAL
[irotaki i=pi-aNt-ak-i=ri
tsamiri]SC
FOC
3m.A=convert-APPL.REAS-PRF-REAL=REL curassow
‘[They said:] ‘Let’s convert him in a curassow, he fights a lot,’ that’s why they
turned him into a curassow.’
4.217
Airorika onimotziro airo opimantavaitzi, ooyana apaniroini naaka, irotaki
akisavakantapintari.
[airo=rika
o=nimo-tz-i=ro
airo
NEG.IRR=COND 3n.m.A=like-EP-REAL=3n.m.O NEG.IRR
o=pimaNta-vai-tz-i
ooy-a=na
apaniroini naaka]MC [irotaki
3n.m.S=sell-DUR-EP-REAL expect-REAL=1SG.O solely
I
FOC
a=kis-av-ak-aNt-apiNt-a=ri]SC
1PL.S=get.upset-REC-PRF-APPL.REAS-HAB-REAL=REL
‘When she doesn’t want to sell [fruit], she expects me solely [to earn money], as a
278
result, we argue.’
4.218 Te anyero ina, ohatatzi oshimata, irotaki kaari onyantana notzimantakarori.
[te
a=ny-e=ro
ina
o=ha-t-atz-i
NEG.REAL 1PL.A=see-IRR=3n.m.O mother 3n.m.S=go-EP-PROG-REAL
o=shima-t-a]MC
[irotaki kaari o=ny-aNt-a=na
3n.m.S=fish-EP-REAL FOC
NEG.P 3n.m.A=see-APPL.REAS-REAL=1SG.O
no=tzimaNt-ak-a=ro=ri]SC
1SG.A=give.birth-PRF-REAL=3n.m.O=REL
‘My mother wasn’t present, she went fishing, that’s why she didn’t see me when I
gave birth.’
(iv) Undesirable possible consequence construction. The dedicated ‘apprehensive’
possible consequence morpheme =kari ‘lest’ is used to mark the subordinate clause
verb. The morpheme =kari always refers to a hypothetical undesirable event described
by the subordinate clause. Examples (4.219-4.220) are quotes from the collected
indigenous traditional advice given by the parents to their children. These admonitions
refer to the crippling fears of untimely death and debilitating illness that Ashéninka
Perené speakers used to possess. To neutralize the threat of an early death, as example
(4.219) shows, the common parenting technique was to wake up the children before
dawn. They were sent to the ice-cold river Perené for a swim and later assigned work to
do to keep them busy. It was believed that more hours of active wake-up time facilitate
longer life of both adults and children. Another extremely unwelcome event was
contracting intestinal parasites or an infectious disease from fellowmen. Example
(4.220) describes one of the steps that should be taken by a guest while visiting a
fellowman in his house, i.e.to make avail of the private facilities as far from the host’s
house as possible.
279
4.219 Ontzimatye pintzinae kapichikitaite, anankoremikari kamanitaantsi, pikamekari
shintsini.
[ontzimatye pi=N-tzin-a-e
kapichikitaite] MC
it’s.necessary 2S=IRR-get.up-REGR-IRR early
[anaNkori-e=mi=kari kamanitaantsi] SC [pi=kam-e=kari
shintsi-ni] SC
cross-IRR=2O=APPR death
2S=die-IRR=APPR be.quick-ADV
‘You have to get up early lest you be in danger of meeting with death, lest you die
quickly.’
4.220 Pihate intyatzini, pitsintakoterikari ashitarori ivanko.
[pi=ha-t-e
intyatzini]MC [pi=tsiNt-ako-t-e=ri=kari
2S=go-EP-IRR other.side
2A=urinate-APPL-EP-IRR=3m.O=APPR
ashi-t-a=ro=ri
i=paNko]SC
have-EP-REAL=3n.m.O=REL 3m.poss=house
‘You should go far lest you urinate where the owner of the house does.’
The single-clause kari-constructions are used as warnings in direct speech,
exemplified by (4.221-4.223).
4.221 Pikaraikari!
pi=kar-a-i=kari
2SG.S=break-REGR-REAL=APPR
‘You’ll break a bone!’
4.222 Piparikari!
pi=pari-i=kari
2SG.S=fall-REAL=APPR
‘You’ll fall!’
4.223 Pipinkikari!
pi=piNki-i=kari
2SG.S=drown-REAL=APPR
‘You’ll drown!’
280
4.4
Other types of clausal linking
Other types of clause linking include (i) contrast, (ii) disjunction, and (iii) manner40.
Each of the above-mentioned clausal construction types will be elaborated on below.
An overview of contrastive, disjunctive, and manner clause linking types is given in
Table 44.
TABLE 44. Contrastive, disjunctive, and manner clause linking devices
Type of clause
linking
Contrast
Disjunction
Manner
Form
Gloss
Function
iro/iroma/irokya
=ha on personal
pronouns
kantzimai(n)tacha
‘in contrast’, ‘but’
‘in contrast’, ‘but’
to contrast direct opposites
to contrast direct opposites
‘nevertheless’,
‘however’
‘or’
‘like’
to contrast speaker’s and
actor’s perspectives
to express an alternative
to indicate a similar manner of
actions performed in both
clauses
to indicate semblance of an
action performed in one clause
with that of another
terika
kimi
shiy
kimi + =rika
‘as if’
(i) Contrast. Contrast refers to the type of clause linking in which the event described by
the first conjunct contrasts with that in the second conjunct. Three contrastive
constructions have been distinguished: the adversative coordinator iro/iroma/irokya
construction ‘but’, ‘in contrast’, e.g. ‘He was sleeping during the day but he was awake
at night’; the contrastive construction with the clausal clitic=ha, e.g. ‘You go to the
woods and bring home a lot of meat, in contrast I never bring anything home’ ; and the
contrastive construction with the counterexpectational coordinator kantzimai(n)tacha
‘nevertheless’, ‘however’, e.g. ‘He told his brother-in-law not to follow him, nevertheless
40
Clause linking in Ashéninka Perené does not formally mark addition. Instead, the language utilizes
juxtaposition of clauses of an equal syntactic status, which can be freely ordered.
281
[unexpectedly] he went’ . Contrast can also be shown by the obsolete personal pronouns
with the -aintsi inflection ‘but I/you/he/she/we’ which express contrastive emphasis on
the participant (see § 3.4.5 for examples and discussion of these archaic pronouns), and
by apposition. Each contrastive construction is addressed in detail below.
The semantic content of the adversative coordinator iro/iroma/irokya has a wide
functional range varying from contrasting direct opposites, as seen in (4.224), to vaguely
signaling a difference between the types of beeds, as seen in (4.225).
4.224
Ari ikantatya imaye irohatzi otsitenitantahya irokya tsitenirikipaye ikantziro
iina...
[ari i=kaNt-aty-a
i=may-e
irohatzi
PP 3m.S=be-PROG-REAL 3m.S=sleep-IRR until
[iro=kya
o=tsiteni-t-aNt-ah-ia]MC
3n.m.S=get.dark-EP-APPL.REAS-REGR-IRR but=EMPH
tsiteniri-ki-paye
night-LOC-PL
i=kaNt-tz-i=ro
i=ina]MC
3m.A=say-EP-REAL=3n.m.O 3m.poss=wife
‘It was the case that he was sleeping until it got dark but during the night he was
saying to his wife…’
4.225
Ari oshaikayetziri incha kitsoiriki, noshintsatantari, iroma iroka chovankeriki,
iroka noshintsatakero, notsanaitakaro.
[ari
o=shaik-a-ye-tz-i=ri
incha kitsoi-ri-ki
there 3n.m.S=be.at-EP-DIST-EP-REAL=REL tree round.thing-NMZ-CL:small.round
no=shintsa-t-aNt-a=ri]MC
[iroma
iroka chovaNkeriki
iroka no=shintsa-t-ak-i=ro
no=tsan-a-it-ak-a=ro]MC
1SG.A=string-EP-APPL.REAS-REAL=REL in.conrast DEM red.&black.seed
DEM 1SG.A=string-EP-PRF-REAL=3n.m.O 1SG.A=hang-EP-ICPL-PRF-REAL=3n.m.O
‘There are many tree seeds there for making strings of adornments, whereas this is
chovankerirki, this I use to string seeds and put it on [necklace made of seeds].’
The contrastive focus clitic =(ma)ha, when it appears in the second conjunct adjoined
to personal pronouns in S/A function (most commonly, with 1SG pronoun naaka ‘I’ and
282
2SG aviroka ‘you’), signals that the actor participant’s activity or state contrasts with that
of the actor participant in the first conjunct. Examples are given in (4.226-4.227).
4.226
Aavakeniri, onkaatakeniri, onaryakeri eentsite, onkivanontahero, onkotsiteniro
oyari, naakaha tekatsi amenenani.
[a-av-ak-e=ni=ri
o=N-kaa-t-ak-e=ni=ri
take-DIR-PRF-IRR=3.O=3m.O 3n.m.A=IRR-bathe-EP-PRF-IRR=3m.O
o=nary-ak-e=ri
eentsi-te
o=N-kivanoNt-ah-e=ro
3n.m.A=lay-PRF-IRR=3m.O 3n.m.poss.child-poss 3n.m.A=wash-REGR-IRR=3n.m.O
o=N-kotsi-t-e=ni=ro
o=ya=ri]MC
3n.m.A=IRR-cook-EP-IRR=3.O=3n.m.O 3n.m.poss=eat=REL
[naaka=ha
I=CNTR
tekatsi
amen-e=na=ni]MC
NEG.EXIST look.after-IRR=1SG.O=REL
‘She [the mother] will receive the baby for her, she will bathe it for her, put it in
bed, wash all the dirty staff, she will cook food for the woman, but I [as an
orphan] have no one who will look after me.’
4.227 Te, iyora iyotziro inyaatsata, naakamaha te noyopirotero.
[te
iyora i=yo-tz-i=ro
i=nyaatsa-t-a]MC
NEG DEM
[naaka=maha
3m.A=know-EP-REAL=3n.m.O 3m.S=play-EP-REAL I=CNTR
te
no=yo-piro-t-e=ro]MC
NEG.REAL 1SG.A=know-AUG-EP-IRR=3n.m.O
‘No, this one knows how to play, but I don’t know [how to play the game].’
Counterexpectational coordinator kantzimai(n)tacha ‘nevertheless’ is used in
situations when the speaker’s and the actor’s perspectives diverge, when the speaker’s
reasonable assumptions about the actor’s activity or state appear to be wrong, often to the
speaker’s surprise. There is no opposition made between the event described in the first
conjunct and that of the second conjunct, as seen in (4.228-4.229).
283
4.228 Okantai ina aisatzi apa, kantzimaitacha maatsi evankaripaye, kaari
kimisantahirori.
[o=kaNt=ai
ina
aisatzi apa]MC [kantzimaitacha tzimatsi
3n.m.A=say=1PL.O mother also
father
nevertheless
EXIST
evaNkari-paye kaari
kimisaNt-ah-i=ro=ri]MC
young.man-PL NEG.P pay.attention-REGR-REAL=3n.m.O=REL
‘Mother told us [this], [and] father too, nevertheless [to my surprise], there are
young people who don’t pay attention [to this advice].’
4.229
Pairani tzimatsi aparoni ashaninka vatsantsi, aintziri iina, kantzimaitacha te
ikovintsate.
[pairani tzimatsi aparoni ashaniNka
vatsantzi a-aintsi=ri
before EXIST one
our.fellowman fat.m
take-STAT.T=REL
[kantzimaitacha te
i=kovintsa-t-e]MC
i= ina]MC
3m.poss= wife nevertheless
NEG.REAL 3m.S=hunt-EP-IRR
‘Long ago there was a fat man who had a wife, nevertheless [surprisingly] he
wasn’t a good hunter.’
(ii) Disjunction. The negative conditional terika ‘if not’ functions as a grammatical
marker of clausal disjunction, e.g. ‘You will marry a dog or you will marry this man.’ In
this type of clausal linking, which involves symmetrical alternatives, the two clauses have
equal status and can occur in either order. Examples are given in (4.330-4.31).
4.330 Nonkaakitakiri irokiki, avisanakiri iito terika nonkaatakiri inchashipatsaini.
[no=N-kaa-ki-t-ak-e=ri
ir=oki-ki
1SG.A=IRR-pour-CL:small.round-EP-PRF-IRR=3m.O 3m.poss=eye-LOC
[te=rika
avis-an-ak-e=ri
i=ito]MC
pass-DIR-PRF-IRR=3m.O 3m.poss=head NEG.REAL=COND
no=N-kaa-t-ak-e=ri
incha-shi-patsa-ini]MC
1SG.A=IRR-bathe-EP-PRF-IRR=3m.O plant-leaf-small.part-DIM
‘I will put pusanga in his eye, and his headache will go away, or I will bathe him
with herbs.’
284
4.331 Paminatye irora kitairiki, samani, sharo, pinkintavaketa, terika paminatye
tsamiri, kintsori, pinkinterita.
[p=amin-aty-e
irora kitairiki samani sharo
2S=see-PROG-IRR DEM peccary paca agouti
pi=N-kiNt-av-ak-e=ta]MC
2S=IRR-shoot.arrows-DIR-PRF-IRR=OPT
[terika p=amin-aty-e
or
2S=see-PROG-IRR
tsamiri kintsori
pi=N-kiNt-e=ri=ta]MC
curassow patridge.sp 2A=IRR-shoot.arrows-IRR=3m.O=OPT
‘You’ll see peccary, paca, agouti, shoot them, or you’ll see curassow, patridge,
shoot them [too].’
(iii) Manner. Grammatical ways of expressing clausal manner linking are limited to the
constructions with verbs kimi ‘be similar, resemble’ and shiy ‘be like’. These verbs mark
‘real’ manner type linking, referring to the similar manner of the actions performed in
both clauses, e.g. ‘The ducks stand in ordered lines like warriors do.’ The verb kimi in
combination with the conditional clitic =rika indicate ‘hypothetical’ manner clause
linking. The ‘hypothetical’ manner type refers to an event described in the adverbial
clause which bears semblance to some action or state portrayed by another clause but
which is unlikely to eventuate in real life, e.g. The ducks stand in ordered lines as if they
were warriors.’ Examples are provided in (4.332-4.334).
4.332 Te inkitatashitahyari okimita ikitaitziri pairani.
[te
i=N-kita-t-ashi-t-it-ah-ia=ri]MC
NEG.REAL 3m.A=IRR-bury-EP-APPL.INT-EP-ICPL-REGR-IRR=3m.O
[o=kimi-t-a
i=kita-it-tz-i=ri
pairani]MC
3n.m.S=be.similar-EP-REAL 3m.A=bury-ICPL-IMP.P-REAL=3m.O before
‘They don’t bury the dead like they buried them before.’
285
4.333 Yantavaitzi yaantari ivarite ivankoki ishiyari koyeranki yaye ivarite imavoshiki.
[y=aNt-a-vai-tz-i
y=a-aNt-a=ri
i=vari-te
3m.S=work-EP-DUR-EP-REAL 3m.S=take-APPL.REAS-REAL=REL 3m.poss=food-poss
i=paNko-ki]MC
[i=shiy-a=ri
koye=raNki y=ay-e
3m.poss=house-LOC 3m.A=be.like-REAL=3m.O ant=ADV.P 3m.S=take-IRR
i=vari-te
i=mavoshi-ki]MC
3m.poss=food-poss 3m.S=nest-LOC
‘He works to get the food home like ants take their food to the nest.’
4.334 Anyiri nihaatsapyaki kisaatsantsanaite katari ikimitaka imitsainkyarika maaroni
yovayerityaranki.
[a=ny-i=ri
nihaa-tsapya-ki kisaa-tsantsana-ite katari]MC
1PL=see-REAL=3m.O river-bank-LOC be.black-wide-AUG wild.duck
[i=kimi-t-ak-a
i=mitsaiNk-ia=rika
maaroni
3m.S=be.similar-EP-PRF-REAL 3m.S=be.in.line-IRR=COND all
y=ovayeri-t-ia=raNki]MC
3m.S=fight-EP-IRR=ADV.P
‘We see black ducks on the river banks as if they were all warriors standing in
lines.’
286
RERERENCES & BIBLIOGRAPHY
Adelaar, Willem, & Pieter Muysken. 2004. The languages of the Andes. Cambridge:
Cambridge University Press.
Aikhenvald, A. 2010. Imperatives and commands. Oxford: Oxford University Press.
Aikhenvald, Alexandra. 2007. Typological distinction in word-formation. In T. Shopen
(ed.), Language typology and syntactic description, Vol. III, Grammatical
categories and the lexicon, 2nd edn, 1-64. Cambridge: Cambridge University
Press.
Aikhenvald, Alexandra. 2003. A grammar of Tariana, from Northwest Amazonia.
Cambridge/New York: Cambridge University Press.
Aikhenvald, Alexandra. 2002. Typological parameters for the study of clitics, with
special reference to Tariana. In R.M.W. Dixon & A. Aikhenvald (eds.), Word: A
Cross-Linguistic Typology, 42-78. Oxford: Oxford University Press.
Aikhenvald, Alexandra. 2000a.Transitivity in Tariana. In R. M. Dixon and A.
Aikhenvald (eds.), Changing valency: Case studies in Transititvity, 145-172.
Cambridge: Cambridge University Press.
Aikhenvald, Alexandra. 2000b. Classifiers. A typology of noun categorization devices.
Oxford: Oxford University Press.
Aikhenvald, Alexandra. 1999. The Arawakan language family. In R.M.W. Dixon and A.
Aikhenvald (eds.), The Amazonian languages, 65-106. Cambridge: Cambridge
University Press.
Aikhenvald, Alexandra. 1998. A grammar of Warekena. In D. Derbyshire and G. Pullum
(eds.), Handbook of Amazonian languages, Vol.4, 215-439. Berlin: Mouton
de Gruyter.
Aikhenvald, A. 1995. Person marking and discourse in North Arawak languages. Studia
Linguistica, 49: 153-95.
Alonso Sutta, Alicia. 2010. Avances en el desarrollo de al lengua escrita de los pueblos
yine, shipibo y ashaninka. Lengua y sociedad 10 (1): 79-85.
Alonso Sutta, Alicia. 2007. El classificador –tha en asháninka del Alto Perené
(Katonkosatí). Actas del V Congreso Nacional de Investigacíones LingüísticoFilológicas. Lima: Universidad Ricardo Palma.
287
Anderson, Janice D. 1988. ‘Tema’: una conjunction en ashéninca. Lengaje y Ciencias,
28(2): 1-9.
Anderson, Janice D. 1983. Dos verbos “adverbiales” del ashéninca. In D. Payne & M.
Ballena (eds.), Estudios lingüísticos de textos ashéninca, 79-90.
Yarinacocha: Instituto Lingüístico de Verano.
Anderson, Ronald J. 2005. Differencias regionales reconocidas por los “campa”.
Revista Latinoamericana de Estudios Lingüísticos 10:111-122.
Anderson, Ronald J. 2000. Ashéninka stories of change. Publications in Sociolinguistics,
4. Dallas: SIL International.
Anderson, Stephen R. 2005. Aspects of the theory of clitics. Oxford: Oxford University
Press.
Andrews, Avery. 2007. Relative clauses. In T. Shopen (ed.), Language typology and
Syntactic Description, Vol. II, Complex constructions, 2nd edn, 206-235.
Cambridge: Cambridge University Press.
Bickel, Balthasar, & Johanna Nichols, 2007. Inflectional morphology. In T. Shopen (ed.),
Language typology and syntactic description, Vol. III, Grammatical categories
and the lexicon, 2nd edn, 169-239. Cambridge: Cambridge University Press.
Black, Andrew. 1991. The phonology of the velar glide in Axininka Campa. Phonology
8(2): 183-217.
Burquest, Donald. 2006. Phonological analysis: Functional approach. 3rd edn. Dallas,
TX: SIL International.
Campbell, Lyle. 1997. American Indian languages: The historical linguistics of Native
America. Oxford: Oxford University Press.
Chunga, S., & A. Torres. 2008. Los pueblos indígenas y el derecho a la educación, dentro
del Programa de Formución de maestros Interculturales Bilingües de la
Amazonia Peruana. Laboratorio de Políticas Públicas: 26.
Citoran, Ioana, & Andrew Nevins. 2008. Introduction. Lingua 118:1900-1905.
Citko, Barbara. 2004. On headed, headless and light-headed relatives. Natural language
and linguistic theory 22: 95-196.
Clark, John, Colin Yallop, & Janet Fletcher. 2007. An introduction to Phonetics and
Phonology. 3rd edn. Malden, MA: Blackwell.
288
Comrie, Bernard. 1989. Language universals and syntactic typology. Syntax and
morphology. 2nd edn. Chicago: The University of Chicago Press.
Comrie, Bernard, & Norval Smith. 1977. Lingua Descriptive Studies. Questionnaire.
Lingua 42:1-71.
Corbett, Greville. 2007. Gender and noun classes. In T. Shopen (ed.), Language typology
and syntactic description, Vol. III, Grammatical categories and the lexicon, 2nd
edn, 241-279. Cambridge: Cambridge University Press.
Cristofaro, Sonia. 2008. Utterance complement clauses. In M. Haspelmath, M.S.
Dryer, D. Gil, &B. Comrie (eds.). The World Atlas of Language Structures
Online. Munich: Max Planck Digital Library, chapter 128. Available online at
http://wals.info/static/descriptions/128/wals_feature_128.pdf. Accessed 04/08/10.
Crowhurst, M., & L. Michael. 2005. Iterative footing and prominence-driven stress in
Nanti (Kampa). Language 81(1): 47-95.
Cushimariano Romano, Rubén, & Richer C. Sebastián Q. 2008. Ñaantsipeta asháninkaki
birakochaki. Diccionario Asháninka-Castellano. Versión preliminar. Available
online at http://www.lengamer.org/publicaciones/diccionarios/.
Accessed 01/20/ 2010.
Cysouw, Michael. 2007. Content interrogatives in Pichis Asheninka. Corpus study and
typological comparison. International Journal of Linguistics 73:133-63.
Danielsen, Swintha. 2007. Baure: An Arawak language of Bolivia (Indigenous languages
of Latin America (ILLA) 6). Leiden: CNWS Publications.
Denevan, William M. 1971. Campa Subsistence in the Gran Pajonal, Eastern Peru.
Geographical Review (New York), 61(4): 496-519.
Dick, Simon, Maria E. Hoffman, Jan R. de Jong, Sie Ing Djiang, Harry Stroomer, &
Lourens de Vries. 1981. On the typology of focus phenomena. In T. Hoekstra, H.
van der Hulst, & M. Morotgat (eds.), Perspectives on Functional Grammar, 4174. Dordrecht: Foris Publications.
Dixon, R.M.W. 2010a. Basic linguistic theory, Vol.2: Grammatical Topics. Oxford:
Oxford University Press.
Dixon, R.M.W. 2010b. The semantics of clause linking in typological perspective. In
R.M.W. Dixon & Alexandra Aikhenvald (eds.), The semantics of clause linking,
1-55. Oxford: Oxford University Press.
289
Dixon, R.M.W. 2004. The Jarawara language of Southern Amazonia. Oxford: Oxford
University Press.
Dixon, R.M.W. 2000. A typology of causatives: form, syntax and meaning. In R.M.W.
Dixon and A. Aikhenvald (eds.), Changing valency: Case studies in transitivity,
30-83. Cambridge: Cambridge University Press.
Dobrushina, Nina, Johan van der Auwera, & Valentin Goussev. 2008. The Optative. In
M. Haspelmath, M.S.Dryer, D. Gil, &B. Comrie (eds.). The World Atlas of
Language Structures Online. Munich: Max Planck Digital Library, chapter 73.
Available online at http://wals.info/static/descriptions/73/ wals_feature_73.pdf.
Accessed 05/08/10.
Dryer, M. 2007. Word order. In T. Shopen (ed.), Language typology and syntactic
description, Vol. I, Clause structure, 2nd edn, 61-130. Cambridge: Cambridge
University Press.
Epps, Patience. 2009. Between headed and headless relative clauses: evidence from Hup.
Paper presented at the Eighth Biannual Conference of the Association for
Linguistic Typology, Berkeley, July 23-26, 2009.
Epps, Patience. 2008. A grammar of Hup. Berlin: Mouton de Gruyter.
Fabre, Alain. 2005. Diccionario etnolingüístico y guía bibliográfica de los pueblos
indígena sudamericanos. Edición electrónica. Available online at
http://word.bienesyautos.com/word-Alain-Fabre-2005--Diccionario/. Accessed
11/03/10.
Falcon, Ccenta Pedro. 1994. Procesos de lenicíon y elision en el ashéninca del Alto
Perené (tesis). Lima: UNMSM.
García Salazar, Gabriela Victoria. 1997. Generalidades de la morfologia y fonologia del
ashéninca del Ucayali. Revista Latinoamericana de estudios etnolingüísticos 9.
Gines, Henry. 2001. Inventario turístico de la Provincia de Chanchamayo. Aspectos
generales. GHEGS American studies.
Gordon, Raymond Jr.(ed.). 2005. Ethnologue: Languages of the World, 15th edn. Dallas:
SIL International.
Grenoble, Lenore, & Lindsay Whaley. 2006. Saving languages: An introduction to
language revitalization. Cambridge: Cambridge University Press.
290
Grinevald, Colette. 2000. A morphosyntactic typology of classifiers. In Gunter Senft
(ed.), Systems of Nominal Classification, 50-92. Cambridge: Cambridge
University Press.
Guillaume, Antoine. 2004. A Grammar of Cavineña, an Amazonian Language of
Northern Bolivia. PhD thesis, Research Centre for Linguistic Typology, La Trobe
University, Melbourne.
Haan, de Ferdinand. 2006a. Irrealis: fact or fiction? Paper presented at the Linguistic
Society of America Meeting, Albuquerque, NM, January 5-8, 2006. Available
online at www.u.arizona.edu/~fdehaan/papers/Irrealis.pdf. Accessed 10/15/2010.
Haan, de Ferdinand. 2006b. Typological approaches to modality. In W. Frawley (ed.),
The expression of modality, 27-70. Berlin/New York: Mouton de Gruyter.
Hale, Kenneth. 1976. The adjoined relative clause in Australia. In R.M.W. Dixon (ed.),
Grammatical categories in Australian languages, 78-105.Canberra: Australian
Institute of Aboriginal Studies.
Halpern, Aaron. 1998. Clitics. In Andrew Spencer & Arnold Zwicky (eds.), Morphology,
101-122. Malden, MA: Blackwell.
Haspelmath, Martin. 2002. Understanding morphology. London/New York: Arnold.
Heise, María, David Payne, Judith Payne, & Elsa Vilchez. 2000. Diccionario escolar
ashaninka/asheninka. Peru: Ministerio de Educacíon.
Hornberger, Nancy. 2000. Education policy and Practice in the Andes: Ideological
Paradox and Intercultural Possibility. Anthropology and Education Quarterly,
31(2):173-201.
Hualde, José Ignacio. 2005. The sounds of Spanish. Cambridge: Cambridge University
Press.
Jacinto Santos, Edwin. 2010. Elautoreconocimiento de la identidad: ashaninka vs.
asheninka. Lengua y sociedad 10(1): 67-78.
Jacinto Santos, Edwin, & Daniel Bernales Quillatupa. 2007. Sankinarentsi. La Merced,
Peru: CIAMB.
Jacinto Santos, Edwin. 2004. Gramática asháninka basica. La Merced, Peru: CIDEARP.
291
Jendraschek, Gerd. 2009. Clause linkage in a language without coordination: the adjoined
clause in Iatmul. In Johannes Helmbrecht, Yoko Nishina, Yong-Min Shin,
Stavros Skopeteas, & Elisabeth Verhoeven (eds.), Form and Function in
Language Research, 139-150. Berlin/New York: Mouton de Gruyter.
Krauss, Michael. 2007. Classification and terminology for degrees of language
endangerment. In Matthias Brenzinger (ed.), Language diversity endangered, 1-8.
Berlin/New York: Mouton de Gruyer.
Ladefoged, Peter. 2005. Vowels and consonants. 2nd edn. Malden, MA: Blackwell.
Ladefoged, Peter. 2003. Phonetic data analysis: An introduction to fieldwork and
instrumental techniques. Malden/Oxford: Blackwell.
Ladefoged, Peter. 2001. A course in phonetics. 4th edn.Boston, MA: Heinle & Heinle.
Ladefoged, Peter, Jenny Ladefoged, & Daniel Everett. 1997. Phonetic structures of
Banawá, an endangered language. Phonetica 1997 (54): 94-111.
Ladefoged, Peter, & Ian Maddieson. 1996. The sounds of the World’s Languages.
Malden, MA: Blackwell.
Lambrecht, Knud. 1994. Information structure and sentence form. Cambridge:
Cambridge University Press.
Lang, Peter. 2004. The grammar of Trio. Frankfurt: Europäishe Verlag der
Wissenschaften.
Lehmann, Christian. 2004. Nexion-Complex Sentences (course materials). Erfurt:
Universität Erfurt.
Michael, Lev. 2010. The semantics of clause linking in Iquito. In R.M.W. Dixon &
Alexandra Aikhenvald (eds.), The semantics of clause linking, 145-166. Oxford:
Oxford University Press.
Michael, Lev. 2009. The Nanti reality status system: Implications for the cross-linguistic
validity of the realis/irrealis contrast. Paper presented at CELCNA, University of
Utah, Salt Lake City, March 28, 2009.
Michael, Lev. 2009. Comparative Kampan vocabulary list. ms.
Michael, Lev. 2008. Nanti evidential practice: Language, knowledge and social action in
an Amazonian society. Ph.D. thesis, University of Texas at Austin.
292
Michael, Lev. 2007. Mood and negation in Nanti (Kampa, Arawak). Available online at
htpp://lev.d.michael.googlepages.com/research_publications. Accessed 06/12/09.
Mihas, Elena. 2010a. Orthography development in Ashéninka Perené (Arawak): a case
study. Cadernos de Etnolingüística 2 (2). Available online at
http:// www.etnolinguistica.org/issue:vol2n2.
Mihas, Elena. 2010b. Nominal classification in Ashéninka. LSA 2010 Annual Meeting
Extended Abstracts. Available online at http://www.elanguage.net/journal
s/index.php/lameeting/index.
Mihas, Elena. 2010c. Forthcoming. Gender system in Ashéninka Perené. UniverSOS,
Revista de Lenguas Indígenas y Universos Culturales 7.
Mithun, Marianne. 2005. Beyond the core: Typological variation in the identification of
participants. International Journal of Linguistics 71: 445-472.
Mithun, Marianne. 1991. Active/agentive case marking and its motivations. Language
67:510- 46.
Nevins, Andrew, & Ioana Citoran. 2008. Phonological representations and the variable
patterning of glides. Lingua 118:1979-1997.
Nichols, J. 1986. Head-marking and dependent- marking grammar. Language 62(1):
56-119.
Nieroda, Janine. 2005. Education for Social Solidarity and Revolution: A Closer Look at
Intercultural Bilingual Education in Peru. Languages, Communities, and
Education, 3: 55-64.
Noonan, Michael. 2007. Complementation. In T. Shopen (ed.), Language typology
and syntactic description, Vol.2. Complex constructions, 52-150. Cambridge:
Cambridge University Press.
Payne, David. 2002. Causatives in Asheninka: the case for associative source. In
M. Shibatani (ed.), The grammar of causation and interpersonal manipulation
485-505. Amsterdam: John Benjamins.
Payne , David. 1991. A classification of Maipuran (Arawakan) languages based on
shared lexical retentions. In D. Derbyshire and G.K. Pullum (eds.), Handbook of
Amazonian languages, Vol. 3, 355-499. Berlin: Mouton de Gruyter.
293
Payne, David. 1984. Activity as the encoding of foregrounding in narrative: a case study
of an Asheninca legend. In Robert E. Longacre (ed.), Theory and application in
processing texts in non-Indo-European languages. Papers in Textlinguistics,
43, 49-91. Hamburg: Buske.
Payne, David. 1983. Notas fonológicas y morfofonémicas sobre el ashéninka del Pichis.
In D. Payne & M. Ballena (eds.), Estudios lingüísticos de textos
ashéninca (campa-arawak preandino), 101-112. Yarinacocha: Ministerio de
educación and Instituto Lingüístico de Verano.
Payne, David. 1981. Phonology and morphology of Axininca Campa. SIL publications in
Linguistics, 66. Dallas: SIL and University of Texas at Arlington.
Payne, David. 1980. Diccionario asheninka-castellano. Yarinacocha: Instituto
Lingüístico de Verano.
Payne, David. 1978. Phonology and morphology of Axininca (Apurucayali Campa).
PhD dissertation. University of Texas at Austin.
Payne, David, & Judith Payne. 2005. The pragmatics of split intransitivity in asheninka.
Revista Latinoamericana de Estudios Ethnolingüísticos 10: 37-56.
Payne, David, Judith Payne, & Jorge Sánchez.1982. Morfología, fonología y fonética del
asheninka del Apurucayali. Serie Lingüística Peruana 18. Lima: Instituto
Lingüístico de Verano.
Payne, Judith. 1990. Asheninka stress patterns. In Amazonian linguistics: Studies in
Lowland South American Languages, Doris L. Payne (ed.), 185-209. Austin:
University of Texas Press.
Payne, Judith. 1989. Lecciones para el aprendizaje del idioma ashéninca. Serie
Lingüística Peruana 28. Yarinacocha: Ministerio de Educación y Instituto
Lingüístico de Verano.
Payne, Judith.1982. Directionals as time referentials in Asheninca. Anthropological
Linguistics 24:325-337.
Payne, Thomas. 1997. Describing morphosyntax. A guide for field linguists.
Cambridge/New York: Cambridge University Press.
Reed, Judy C., & David Payne. 1986. Ashéninca (Campa) pronominals. In Ursula
Wiesemann (ed.), Pronominal systems, 323-31.Tübingen: Gunter Narr.
Ribeiro, Darcy & Mary Ruth Wise. 1978. Los grupos etnicos de la Amazonía Peruana.
Lima: Instituto Lingüístico de Verano.
294
Rice, Karen. 1990. Grammar of Slave. Berlin/New York: Mouton de Gruyter.
Romani, Miranda Maggie. 2003. Toponimia en el Gran Pajonal: un studio sobre la
morfologia de los topónimos de filiación ashéninka. Actas del V Congreso
Nacional de Investigaciónes Lingüístico-Filológicas. Lima: Universidad Ricardo
Palma.
Rubino, Carl. 2008. Reduplication. In M. Haspelmath, M.S. Dryer, D. Gil, &B. Comrie
(eds.). The World Atlas of Language Structures Online. Munich: Max Planck
Digital Library, chapter 21. Available online at
http://wals.info/static/descriptions/21/wals_feature_21.pdf. Accessed 10/12/10.
Santos-Granero, Fernando. 2004. The enemy within: Child Sorcery, Revolution, and the
Evils of Modernization in Eastern Peru. In Neil L.Whitehead & Robin Wright, In
Darkness and Secrecy, 272-305. Durham & London: Duke University Press.
Santos-Granero, Fernando. 2002. Saint Christopher in the Amazon: Child Sorcery,
Colonialism, and Violence among the Southern Arawak. Ethnohistory 49 (3):
507-543.
Solís Fonseca, Gustavo. 2003. Lenguas en la Amazonia Peruana. Lima, Peru: Visual
Service S.R.L.
Spring, Cari. 1992. The velar glide in Axininca Campa. Phonology 9(2): 329-52.
Stassen, Leon. 2008. Comparative constructions. In M. Haspelmath, M.S. Dryer, D. Gil,
&B. Comrie (eds.). The World Atlas of Language Structures Online. Munich:
Max Planck Digital Library, chapter 121. Available online at
http://wals.info/static/descriptions/121/wals_feature_121.pdf. Accessed 10/30/10.
Swift, Kenneth. 2008. Morfología del Caquinte (Arawak preandino). 2nd edn. Lima:
Instituto Lingüístico del Verano.
Thomason, Sarah. 2001. Language contact. Washington, D.C.: Georgetown University
Press.
Trudell, Barbara. 1995. Más allá del aula bilingüe alfabetización en las comunidades de
la Amazonia peruana. Comunidades y culturas peruanas, No. 26. Yarinacocha:
Ministerio de Educación y Instituto Lingüístico de Verano.
Van Valin Jr., Robert, & Randy J. La Polla. 1997. Syntax. Structure, meaning, and
function. Cambridge: Cambridge University Press.
Varese, Stefano. 2002. Salt of the mountain. Campa Asháninka history and resistance in
the Peruvian jungle. Norman: University of Oklahoma Press.
295
Weiss, Gerald. 1975. The world of a forest tribe in South America. Anthropological
Papers of the American Museum of Natural History LII/5.
Weiss, Gerald. 1974. Campa organization. American Ethnologist 1(2): 379-403.
Wise, Mary Ruth. 2002. Applicative suffixes in Peruvian Amazonian languages. In
Mily Crevels, Simon van de Kerke, Sérgio Meira, & Hein van der Voort
(eds.), Current studies on South American Indian languages: Selected papers
from the 50th International Congress of Americanists in Warsaw and the Spinoza
workshop on Amerindian languages in Leiden, 2000, 329-244. ILLA, 3. Leiden:
Research School of Asian, African, and Amerindian studies.
Wise, Mary Ruth. 1990. Valence-changing affixes in Maipuran Arawakan languages. In
Doris L. Payne (ed.), Amazonian linguistics: Studies in Lowland South American
Languages, 89-116. Austin: University of Texas Press.
Wise, Mary Ruth. 1986. Grammatical characteristics of Preandine Arawakan languages
of Peru. In D. Derbyshire & G.K. Pullum (eds.), Handbook of Amazonian
languages, 567-642. Berlin: Mouton de Gruyter.
Wise, Mary Ruth. 1985. Indigenous languages of lowland Peru: History and current
status. In H. Manelis Klein & L.R. Stark (eds.), South American Indian
Languages, 194-221. Austin: University of Texas Press.
296
APPENDIX A
Basic vocabulary list
1.
2.
3.
4.
5.
I
you (SG)
he
she
we
6.
7.
you (PL)
they
8.
this
9.
that
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
that over
there
here
there
over there
who
what
where
when
how
not
20.
21.
22.
23.
all
many
some
few
24.
25.
26.
27.
28.
other
one
two
three
big
naaka
aviro(ka)
iri(ro)
iroo
naakaite (excl),
arokaite (incl)
avirokaite
iriroite (m),
irooite (n.m.)
iyoka (m), iroka
(n.m.)
iyora (m),irora
(n.m.)
yoNta (m),
iroNta (n.m.)
aka~ haka
ara~hara
aNto
niNka
tsika paita
tsika
tsika ikaratzi
tsika okaNta
te (REAL), airo
(IRR)
maaroni
osheki
kapicheeni
kapicheeni
pashini
aparoni
apite
mava
aNtari (m), aNtaro
(n.m.)
29.
30.
31.
32.
33.
long
wide
thick
heavy
small
santsari
aNtarosaNte
tsateNto
tenari(ni)
iyaani(ni)
34.
35.
short
narrow
36.
thin
37.
woman
iyaani
irihani;
manchaki(ni)
matsari (m), matsaro
(n.m.)
kooya
38.
man
shiraNpari
39.
40.
41.
42.
43.
44.
45.
46.
47.
to think
to smell
to fear
to sleep
to live
to die
to kill
to fight
to hunt
-kiNkishiri-;-shiri-asaNk-tsarov
-ma-any-kam-v-mana-; -vayeri- komintsa-
48.
49.
50.
51.
to hit
to cut
to split
to stab
(pierce with
an arrow)
to scratch
to dig
to swim
to fly
to walk
-pos-chek-viri-;-tsipary-kiNt-
52.
53.
54.
55.
56.
-kitsari-kiy-ama-ar-kin-; anii-; -avis-
297
57.
58.
59.
60.
61.
62.
63.
64.
65.
to come
to lie (down)
to sit
to stand
to turn
to fall
to give
to hold
to squeeze
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
80.
81.
to rub
to wash
to wipe
to pull
to push
to throw
to tie
to sew
person
child
wife
husband
mother
father
animal
fish
82.
83.
84.
85.
86.
87.
88.
89.
90.
91.
92.
93.
94.
bird
dog
louse
snake
worm
tree
forest
stick
fruit
seed
leaf
root
bark (of a
tree)
flower
grass
95.
96.
-pok-nari-saik-katziy-shoNk-; -pitsok-pari-p-airik-kaviriNka-;
-pichov-tsiriNk-kiv-; -kivanoNt-shet-~-shit-noshik-otyaNk-; -shiriNk-ook-oisot-shitzikatziri
eentsi
-ina
-ime~-imi
ina; -niro
apa;-iri
piratsi
shima
97.
98.
99.
100.
101.
102.
103.
104.
105.
rope
skin
meat
blood
bone
fat
egg
horn
tail
shivitsa
-mishina
vatsatsi
-iraani
-toNki
-yeNka
-itsoki
-ichee
-tziNko
106.
107.
108.
109.
110.
111.
112.
113.
114.
115.
116.
117.
118.
119.
120.
121.
feather
hair
head
ear
eye
nose
mouth
tooth
tongue
fingernail
foot
leg
knee
hand
wing
belly
tsimiri
otsitzi
netsi
maraNki
chapitsi
inchato
aNtami
inchakii
chochoki
-itsoki
-ishi; tsipana
paritsa
inchataki
122.
123.
124.
125.
126.
127.
128.
129.
130.
131.
132.
133.
134.
guts
neck
back
breast
heart
liver
to drink
to eat
to bite
to suck
to spit
to vomit
to blow
-shivaNki
-ishi
-ito
-yeNpita
-oki
-kirimashi;-kiriki
-pa(a)Nte
-aiki
-nene
-shetaki
-itzi
-tavaato
-yerito
-ako
-shivaNki
-motya;
-tsomoNte
-sheto
-kentsi
-taapi~-tapi
-tsomi;-nee
-saNkani
-irapana
-ir-ov-~-v-atsik-tsot-choNt-; -tyoNt-kamaraNk-tasoNk-
inchatyaki
savoro
135.
136.
to breathe
to laugh
-aniNk-shiroNt-
298
137.
138.
139.
140.
141.
142.
143.
144.
145.
146.
to see
to hear
to know
to count
to say
to sing
to play
to float
to flow
to freeze
-ny-; -amin-kim-yo-ookot-kaNt-pantsa-nyaatsa-ama-shiy-katsiNkai-
178.
179.
180.
181.
182.
183.
184.
185.
186.
187.
white
black
night
day
year
warm
cold
full
new
old
147.
148.
149.
150.
151.
152.
153.
154.
155.
156.
157.
158.
159.
160.
161.
to swell
sun
moon
star
water
rain
river
lake
sea
salt
stone
sand
dust
earth
cloud
-anoNkooryaatsiri
kashiri
iNpookiro
nihaa
iNkani
pareni
iNkaari
iNkaari
tzivi
mapi
iNpaneki
samaNpo
kipatsi
miNkori
188.
189.
190.
191.
192.
193.
194.
195.
196.
197.
198.
199.
200.
201.
202.
good
bad
rotten
dirty
straight
round
sharp
dull
smooth
wet
dry
correct
near
far
right
162.
163.
164.
165.
166.
167.
168.
169.
170.
171.
172.
173.
174.
175.
176.
177.
to drizzle
sky
wind
rainbow
snow
ice
fog
smoke
fire
ash
to burn
road (trail)
mountain (hill)
red
green
yellow
-pariNkaiNkite
taNpyaa
parinkari;oyichari
kachaari
paamari
samaNpo
-taavotsi
otzishi
kityoNkari
natsiyari
kiteriri
203.
204.
205.
206.
207.
208.
209.
210.
left
at
in
with
and
if
because
name
kitamarori
kisaari; potsitari
tsiteniri
kitaiteri
osarentsi
katsiriNkari
katsiNkari
haikiri(ni)
ovakerari
kiNkivari/ro;
kiNkivarotaki
kameetsari
kaaripirori
patsaaki
kipatsiri
taNpatsikari
-ki-; -na(ki)tsaiNpiri
te otsaiNpite
karini
asheri
piryari
iripirori
okaakini
iNtaina
nakohanori;
nakopirori
naNpate
-ki
-ki
itsipatairo/ri; -aNt
i/oponya
arika, aririka
tema, kama
-vairo
299
APPENDIX B
Text samples
B.1
Narrative Nantavairi ‘My job’ by Gregorio Santos Pérez (Villa Perené)
1.
Ironyaka nonkinkitsatavake okaratzi nantapintziri kitaiterikipaye onkantya
namantantavaityari novarite namiri novankoki novakaiyariri neentsitepatsaini.
ironyaaka no=N-kiNkitsa-t-av-ak-e
o=kara-tz-i
now
1SG.S=IRR-tell-EP-DIR-PRF-IRR 3n.m.S=COP.C-EP-REAL
n=aNt-apiNt-tz-i=ri
kitaiteri-ki-paye o=N-kaNt-ia
1SG.S=do-HAB-EP-REAL=REL day-LOC-PL
n=amaNt-aNt-av-a-it-ia=ri
3n.m.S=happen-IRR
no=vari-te
1SG.S=buy-APPL.REAS-DIR-EP-ICPL-IRR=REL 1SG.poss=food-poss
n=am-i=ri
no=paNko-ki
1SG.A=bring-REAL=REL 1SG.poss=house-LOC
no=v-ak-aiy-a=ri=ri
n=eentsi-te-patsaini
1SG.A=eat-CAUS-PL-REAL=3m.O=REL 1SG.poss=child-poss-DIM
‘Now I will tell you all that I do during [all] days in order to buy food that I bring
to my house to feed my children.’
2.
Nintavakyaro naakaha, noyomitantzi eshkueraki, noyomitairi eentsitepataini.
n=iNt-av-ak-ia=ro
naaka=ha no=yomit-an-tz-i
1SG.A=begin-DIR-PRF-IRR=3n.m.O I=EMPH
eskuera-ki
no=yomit-a-i=ri
1SG.S=teach-DIR-EP-REAL
eentsi-patsaini
school-LOC 1SG.A=teach-REGR-REAL=3m.O child-DIM
‘I will begin with me: I teach at school, I teach little children.’
3.
Arika ontzime nante iroka ikantaitziri matematika, maatsi ivapeete.
arika o=N-tzim-e
n=aNt-e
iroka i=kaNt-ai-tz-i=ri
when 3n.m.S=IRR-have-IRR 1SG.S=do-IRR DEM 3m.A=say-IMP.P-EP-REAL=3m.O
matematika tzimtsi i=vapee-te
math
EXIST
3m.poss=paper-poss
‘When I have to do this, they call it math, they have notebooks.’
4.
Noyomitairi nosankinatakairi.
no=yomit-a-i=ri
no=saNkina-t-ak-a-e=ri
1SG.A=teach-REGR-REAL=3m.O 1SG.A=write-EP-CAUS-REGR-IRR=3m.O
‘I teach them, I made them write.’
300
5.
Ponyashitaka arika eentsira iyotanakirorika, yantamaintyaro ivapeeteki irirori
isankinate.
ponyashitaka arika eentsi=ra
i=yo-t-an-ak-e=ro=rika
afterwards when child=DEM 3m.A=know-EP-DIR-PRF-IRR=3n.m.O=COND
y=aNt-amaiNt-ia=ro
i=vapee-te-ki
irirori
3m.A=do-INCH-IRR=3m.O
3m.poss=paper-poss-LOC he
‘Then when the child knows [how to write], he begins doing it in his notebook.’
6.
Nookanakiniri irora ejercicio ikantaitziro iroori antavairintsi, yantavaiteta
eentsipatsaini.
n=ook-an-ak-i=ni=ri
irora ejercicio
1SG.A=leave-DIR-PRF-REAL=3O=3m.O DEM exercise
i=kaNt-ai-tz-i=ro
iroori
3m.A=say-IMP.P-EP-REAL=3n.m.O they.n.m
aNtavai-rintsi
work-NMZ
y=aNt-a-vai-t-e=ta
eentsi-patsaini
3m.S=make-EP-DUR-EP-IRR=OPT
child-DIM
‘I assign exercises to them what they call ‘assignment’; the children should do [it]
for a while.’
7.
Arika intsonkakiroha aparopaye, namineniri yantakirorika kametsa.
arika i=N-tsoNk-ak-e=ro=ha
aparo-paye
when 3m.A=IRR-finish-PRF-IRR=3n.m.=EMPH one-PL
n=amin-e=ni=ri
y=aNt-ak-e=ro=rika
kametsa
1SG.A=examine-IRR=3O=3m.O 3m.A=make-PRF-IRR=3n.m.O=COND well
‘When each of them finishes it [the assignment], I’ll inspect it for them if they did
it well.’
8.
Maatsi pashinipaye, kaari tsonkironi, yookashityaro, airo yantzirotzi.
tzimatsi pashini-paye kaari
tsoNk-e=ro=ni
EXIST other-PL
NEG.P finish-IRR=3n.m.O=REL
y=ook-ashi-t-ia=ro
airo
y=aNt-tz-i=ro~tzi
3m.A=leave-APPL.INT-EP-IRR=3n.m.O NEG.IRR 3m.A=make-REAL=3m.O~EMPH
‘There are others, that those who will not finish it will be the case, they will
abandon it, they just won’t do it.’
301
9.
Airorika yantziroha, ontzimatye nompavaheri pashini, nonkantavaheri onkantya
iyoperotantanakyarori.
airo=rika
y=aNt-tz-i=ro=ha
oNtzimatye
NEG.IRR=COND 3m.A=make-EP-REAL=EMPH be.necessary
no=N-p-av-ah-e=ri
pashini no=N-kaNt-av-ah-e=ri
1SG.A=IRR-give-DIR-REGR-IRR=3m.O other
1SG.A=IRR-say-DIR-REGR-IRR=3m.O
oNkaNtya i=yo-pero-t-aNt-an-ak-ia=ro=ri
so.that
3m.A=learn-AUG-EP-APPL.REAS-DIR-PRF-IRR=3n.m.O=REL
‘If they don’t do it, I have to give them another assignment, I will explain [it] to
them so that they could learn it better.’
10.
Itsonkantyarori pashini, arika yantero yantavairi.
i=tsoNk-aNt-ia=ro=ri
pashini arika
3m.A=finish-APPL.REAS-IRR=3m.O=REL other when
y=aNt-e=ro
y=aNtavai-ri
3m.A=make-IRR=3n.m.O 3m.poss=work-NMZ
‘So that they finish another [task], when they do their assignment.’
11.
Maatsi aisatzita eentsipaye, kaari tsonkapiroteroni.
tzimatsi aisatzi=ta eentsi-paye kaari
tsoNk-a-pero-t-e=ro=ni
EXIST also=OPT child-PL
NEG.P finish-EP=AUG-EP-IRR=3n.m.O=REL
‘There are also [other] children; that those who will finish it [the assignment] will
not be the case.’
12.
Isaikashivaita, intaani inyaatsavaitya.
i=saik-ashi-vai-t-a
iNtaani i=nyaatsa-vai-t-ia
3m.S=be.at-APPL.INT-DUR-EP-REAL only
3m.S=play-DUR-EP-IRR
‘They are just sitting, they only will play.
13.
Arika ankantavitakyari, ‘Airo pinyaatsavaita, airo pinyaatsavaita,’ airo tekatsi
inkante, irohatzi yaminashitya sontoki, airo yantzirotzi yantavairi.
arika a=N-kaNt-a-vi-t-ak-ia=ri
airo
pi=nyaatsa-vai-t-a
when 1PL.A=IRR-say-EP-FRUS-PRF-IRR=3m.O NEG.IRR 2S=play-DUR-IRR-EP-REAL
airo
pi=nyaatsa-vai-t-a
airo
tekatsi
i=N-kaNt-e
NEG.IRR 2S=play-DUR-IRR-EP-REAL NEG.IRR NEG.EXIST 3m.S=IRR-say-IRR
irohatzi y=amin-ashi-t-ia
soNtoki
airo
until
3m.S=look-APPL.INT-EP-IRR wide.open.eye NEG.IRR
y=aNt-tz-i=ro~tzi
y=aNtavai-ri
3m.A=make-REAL=3n.m.O~EMPH 3m.poss=work-NMZ
‘Whe we tell them, ‘Don’t play, don’t play,’ they won’t say anything, only will
302
look [at you] with big eyes, [but] they won’t do the assignment.’
14.
Aisatzi arika inkisakakai, ontzimatye ankaimakanteri iriri, paitarika
yantapainteri, onkantyakya maaroni avitsikahero kameetsa, yantakirira eentsi.
aisatzi arika i=N-kis-ak-ak=ai
oNtzimatye
also when 3m.A=IRR-be.angry-CAUS-PRF=1PL.O be.necessary
a=N-kaim-ak-aNt-e=ri
ir=iri
paita=rika
1PL.A=IRR-call-PRF-APPL.REAS-IRR=REL 3m.poss=father WH=COND
y=aNt-apaiNt-e=ri
oNkaNtia=kya maaroni a=vitsik-ah-e=ro
3m.S=make-once-IRR=REL so.that=EMPH all
1PL.A=fix-REGR-IRR=3n.m.O
‘Also when they make us [teachers] angry, we have to call their father, so that all,
whatever he [the child] did, could be resolved.’
15.
Arika imposavakakya o intsitonkakero aparoni kooya, onkamantahero iniro.
arika i=N-pos-av-ak-ak-ia
o i=N-tsitoNk-ak-e=ro
when 3m.S=IRR-hit-RCP-CAUS-PRF-IRR or 3m.A=IRR-hit-PRF-IRR=3n.m.O
aparoni kooya i=N-kamaNt-ah-e=ro
i=niro
one
woman 3m.A=IRR-inform-REGR-IRR=3n.m.O 3m.poss=mother
‘When they fight each other or hit a woman, he [the child] will notify his mother
about it.’
16.
Ompokahe iniro, naakataki vitsikaironi nyaantsi.
arika o=N-pok-ah-e
i=niro
naakataki
when 3n.m.S=IRR-come-REGR-IRR 3m.poss=mother I.FOC
vitsik-a-e=ro=ni
fix-REGR-IRR=3n.m.O=REL
‘When his mother comes, I will be the one who will resolve it [the issue].’
17.
Irotaki aisatzita arika noyomitairi nonyaanate aparonira, ironyaaka noyomitairi
inyaanira, ikantziri virakocha, irotaki naparoninkari.
irotaki aisatzi=ta arika no=yomit-a-e=ri
no=nyaana-t-e
FOC also=OPT when 1SG.A=teach-REGR-IRR=3m.O 1SG.S=instruct-EP-IRR
aparoni=ra ironyaaka no=yomit-a-e=ri
one=DEM now
i=nyaa-ni=ra
1SG.A=teach-REGR-IRR=3m.O 3m.poss=word-poss=DEM
i=kaNt-tz-i
virakocha irotaki n=aparoni-Nka-ri
3m.S=say-EP-REAL
outsider FOC 1SG.poss=one-NMZ-NMZ
‘Also when I teach them, I will speak their language, I will teach in his language,
they say, the language of ‘virakocha’ (of outsiders), in only one language.’
303
18.
Ovakeraini ovakira noyomitakiri eentsipaye anta, noyomitairi inyaaniki,
nonyaaniki naari.
ovakeraini ovakera no=yomit-ak-i=ri
eentsi-paye aNta
recently when 1SG.A=teach-PRF-REAL=3m.O child-PL over.there
no=yomit-a-i=ri
i=nyaa-ni-ki
no=nyaa-ni-ki
naari
1SG.A=teach-REGR-REAL=3m.O 3m.poss=word-poss-LOC 1SG.poss=word-poss-LOC I
‘When recently I taught children over there, I taught them in their language,
[which is] my language too.’
19.
Noyomitaitzitavakyari nonyaanikinta, iyotzirinta eentsi.
no=yomit-a-itz-it-av-ak-ia=ri
no=nyaa-ni-ki=Nta
1SG.A=teach-EP-ICPL-before-DIR-PRF-IRR=3m.O 1SG.poss=word-poss-LOC=DEM
i=yo-tz-i=ri=Nta
eentsi
3m.S=know-EP-REAL=3m.O=ADV child
‘First I will teach in my language, which the child knows.’
20.
Ponyashitakaha noyomitahiri, novakotahiri inyaanikinta virakocha.
ponyashitaka=ha
no=yomit-ah-e=ri
afterwards=EMPH 1SG.A=teach-REGR-IRR=3m.O
no=v-ako-t-ah-i=ri
i=nyaa-ni-ki=Nta
virakocha
3m.poss=word-poss-LOC=DEM outsider
‘Then I will teach them, I will translate it into the Spaniards’ language.’
1SG.A=place-APPL-EP-REGR-REAL=3m.O
21.
Ari nokantzirori pairani, ovakira nintakaro noyomitairi eentsipaye saikatsiri
Pirimorishiki (Koviryaki).
ari no=kaNt-tz-i=ro=ri
pairani ovakira
PP 1SG.A=do-EP-REAL=3n.m.O=REL before when
n=iNt-ak-a=ro
no=yomit-a-i=ri
1SG.A=begin-PRF-REAL=3n.m.O 1SG.A=teach-REGR-REAL=3m.O
eentsi-paye saik-atsi=ri
Pirimorishi-ki
child-PL
be.at-STAT=REL Puerto.Bermudez
‘This is what I did long ago when I began teaching children who lived in Puerto
Bermudez.’
304
22.
Nintakari saikatsiri aka ikantaitzi Ooheroniki, ironyaaka ikantaitahiro San
Miguel Centro Marankiari.
n=iNt-ak-a=ri
saik-atsi=ri
aka i=kaNt-a-it-tz-i
1SG.A=begin-PRF-REAL=3m.O be.at-STAT=REL here 3m.S=say-before-EP-REAL
Ooheroni-ki ironyaaka i=kaNt-ai-tz-i=ro
Fatigue-LOC now
3m.A=say-IMP.P-EP-REAL=3n.m.O
San Miguel Centro Marankiari
village’s.name
‘I began [to work] with the children who lived here [in the village] called
Ooheroni (Fatigue), now it is called San Miguel Centro Marankiari.’
23.
Ironyaaka nonkamantemi pashini tsika paita naantari pashini kiriiki novantari.
ironyaaka no=N-kamaNt-e=mi
pashini tsika paita
now
1SG.A=IRR-inform-IRR=2O other WH
n=a-aNt-a=ri
pashini kiriiki
1SG.S=take-APPL.REAS-REAL=REL other money
no=v-aNt-a=ri
1SG.S=eat-APPL.REAS-REAL=REL
‘Now I will tell you another [story] how I obtain money to eat [buy food].’
24.
Aisatzita noyotzirora nonyaatsata, noyotakiro aisatzita tsika okanta yaminaitzika
ikantya inyaatsatantyari atziripaye kametsa.
aisatzi=ta no=yo-tz-i=ro=ra
no=nyaatsa-t-a
also=OPT 1SG.A=know-EP-REAL=3n.m.O=ADV 1SG.S=play-EP-REAL
no=yo-t-ak-i=ro
aisatzi=ta tsika o=kaNt-a
1SG.A=know-EP-PRF-REAL=3n.m.O also=OPT WH 3n.m.S=happen-REAL
y=amin-a-it-tz-i=ka
i=kaNt-ia
3m.S=examine-EP-ICPL-EP-REAL=ADV 3m.S=do-IRR
i=nyaatsa-t-aNt-ia=ri
atziri-paye kametsa
3m.S=play-APPL.REAS-EP-IRR=REL person-PL well
‘Also when [because] I know how to play it [the game], I also know how to assess
if people play the game well.
25.
Ikantziri virakocha arbitro.
i=kaNt-tz-i=ri
virakocha arbitro
3m.A=say-EP-REAL=3m.O outsider arbiter
‘They call it in Spanish ‘arbiter’.
305
26.
Aisatzi naari ipinaitana.
aisatzi naari i=pina-ai-t-a=na
also
I
3m.A=pay-IMP.P-EP-REAL=1SG.O
‘They also pay me.’
27.
Okimitatya savararokiranki nohatatitzi tonkarikinta Metaarokiha, ari
nomoNkaratziri inyaatsavaita virakochapaye.
okimitatya savaro-ki=raNki
no=ha-t-a-it-tz-i
for.example Saturday-LOC=ADV.P 1SG.S=go-EP-REP-ICPL-EP-REAL
toNkari-ki=Nta
Metaaro-ki=Nta
ari no=moNkara-tz-i=ri
hilltop-LOC=DEM village’s.name-LOC=DEM PP 1SG.A=measure-EP-REAL=3m.O
i=nyaatsa-vai-t-a
virakocha-paye
3m.S=play-DUR-EP-REAL outsider-PL
‘For example, yesterday, on Saturday, I went up there to Metraro to judge
the games of mestizo players.’
28.
Ipinaitakana, namahi kapichiini kiriiki, nopapahiro noina irotaki
okamitantantyari avahyari.
i=pina-ai-t-ak-a=na
n=am-ah-i
kapichiini
3m.A=pay-IMP.P-EP-PRF-REAL=1SG.O 1SG.S=bring-REGR-REAL some
kiriiki no=p-ap-ah-i=ro
no=ina
irotaki
money 1SG.A=give-DIR-REGR-REAL=3n.m.O 1SG.poss=wife FOC
o=kamitaNt-aNt-ia=ri
a=v-ah-ia=ri
3m.S=buy-APPL.INST-IRR=3m.O 1PL.S=eat-REGR-IRR=REL
‘They paid me, I brought home some money, I gave it to my wife, with it she will
buy something to eat.’
29.
Oshetehi tominkokiranki nohataitzi aka Chirani ikantaitziro Tsirani.
oshetehi tomiNko-ki=raNki
no=ha-t-a-it-tz-i
aka
yesterday Sunday-LOC=ADV.P 1SG.S=go-EP-REP-ICPL-EP-REAL here
Chirani
i=kaNt-ai-tz-i=ro
Tsirani
village’s.name 3m.A=say-IMP.P-EP-REAL=3n.m.O village’s.name
‘Yesterday on Sunday I went to Chirani called Tsirani [in our language].
30.
Noshiyakotanaka nomoteteki, nookanakiro intakironta, intatzikironta.
no=shiy-ako-t-an-ak-a
no=mote-te-ki
1SG.S=run-APPL-EP-DIR-PRF-REAL 1SG.poss=bike-poss-LOC
n=ook-an-ak-i=ro
iNtakiroNta intatzikiroNta
1SG.A=leave-DIR-PRF-REAL=3n.m.O outside
other.side.of.river
‘I rode my bike and left it outside, on the other side of the river.’
306
31.
Ari ironyaaka aparoni vaaro, ari ironyaaka anoshikakotyari intatzikironta, ari
imontyakina.
ari ironyaaka aparoni vaaro
ari ironyaaka a=noshik-ako-t-ia=ri
PP now
one
cable.car PP now
1PL.A=pull-APPL-EP-IRR=3m.O
iNtatzikiroNta
ari i=moNty-ak-i=na
other.side.of.river PP 3m.A=cross-CAUS.SOC-REAL=1SG.O
‘There is a cable car, we should push it to the other side, and they [the cable
service workers] crossed with us.’
32.
Ari nokinavaitzi, noyotziro ironyaaka onyaatsari kooya ikantaitziri ‘voli’, aisatzi
noyotziro noyomarotziro.
ari no=kin-a-vai-tz-i
no=yo-tz-i=ro
ironyaaka
PP 1SG.S=walk-EP-DUR-EP-REAL 1SG.A=know-EP-REAL=3n.m.O now
o=nyaatsa-ri
3n.m.poss=play-NMZ
kooya i=kaNt-ai-tz-i=ri
voli
woman 3m.A=say-IMP.P-EP-REAL=3m.O volleyball
aisatzi no=yo-tz-i=ro
no=shoiniNk-ako-tz-i=ro
also
1SG.A=know-EP-REAL=3n.m.O 1SG.A=whistle-APPL-EP-REAL=3n.m.O
‘I was walking for a while, I know [how to play] the woman’s game called
volleyball; I also know how to judge it.’
33.
Tsika ikantakantaitziro onyaatsatantyari kametsa.
tsika i=kaNt-ak-aNt-a-it-tz-i=ro
WH 3m.A=do-CAUS-APPL.REAS-EP-ICPL-EP-REAL=3m.O
o=nyaatsa-t-aNt-ia=ri
kametsa
3n.m.S=play-EP-APPL.REAS-IRR=REL well
‘How to make it for them to play well’.
34.
Irotaki aisatzita naaka ipinaitanari.
irotaki aisatzi=ta naaka i=pina-ai-t-a=na=ri
FOC also=OPT I
3m.A=pay-IMP.P-EP-REAL=1SG.O=3m.O
‘For this they pay me.’
35.
Maatsi irora notzivanaporokite tonkarikintaha.
tzimatsi irora no=tzivana-poroki-te
toNkari-ki=Nta=ha
EXIST DEM 1SG.poss=pineapple-group-poss hilltop-LOC=DEM=EMPH
‘I have a pineapple orchard atop the hill over there.’
307
36.
Ironyaakama atsamaitatziro, ahatzi, aapintzi kapichiini kiriiki onkantya
avantyari.
ironyaaka=ma a=tsamai-t-atz-i=ro
ahatzi a-apiNt-tz-i
now=DUB
1PL.A=clean.up-EP-PROG-REAL=3n.m.O also take-HAB-EP-REAL
kapichiini kiriiki oNkaNtya a=v-aNt-ia=ri
little
money so.that
1PL.S=eat-APPL.REAS-IRR=REL
‘Right now we are cleaning it, also we get some money [from sale of pineapple]
so that we could eat.’
37.
Arika intzimanahe, aitaki ampavakahya maaroni.
arika i=N-tzim-an-ah-e
aitaki a=N-p-av-ak-ah-ia
maaroni
when 3m.S=have-DIR-REGR-IRR PP 1PL.S=IRR-give-REC-PRF-REGR-IRR all
‘When there is money, we share between all [members of the family].’
38.
Airorika intzimi, ontzimatye amine antavairintsi.
airo=rika
i=N-tzim-e
oNtzimatye amin-e
aNtavai-rintsi
NEG.IRR=COND 3m.S=IRR-have-REAL be.necessary look.for-IRR work-NMZ
‘When there is no money, we have to look for work.’
39.
Ironyaaka irora notsipatari opimantavaitzi kapichiini aisatzita.
ironyaaka irora no=tsipa-t-a=ri
o=pimaNt-a-vai-tz-i
now
DEM 1SGS=join-EP-REAL=REL 3n.m.S=sell-EP-DUR-EP-REAL
kapichiini aisatzi=ta
little
also=OPT
‘My wife sells [fruit] a little bit too.’
40.
Airorika onimotziro, airo opimantavaitzi, ooyana apaniroini naaka, irotaki aisatzi
akisavakantapintari.
airo=rika
o=nimo-tz-i=ro
airo
NEG.IRR=COND 3n.m.A=like-EP-REAL=3n.m.O NEG.IRR
o=pimaNt-a-vai-tz-i
ooy-a=na
apaniroini naaka irotaki
3n.m.S=sell-EP-DUR-EP-REAL expect-REAL=1SG.O solo
I
FOC
aisatzi a=kis-av-ak-aNt-apiNt-a=ri
also 1PL.S=get.angry-RCP-CAUS.SOC-APPL.REAS-HAB-REAL=REL
‘When she doesn’t like it, she won’t sell [fruit], expects me only [to work], that’s
why we argue with each other.’
308
41.
‘Pantavaitahaitetyami avirori, naaka apaniroini antavaitatsini, naaka amatsini,
iroka naaka kamitantatsini, iroka naakataki.’
p=aNtavai-t-ah-a-it-e=tya=mi
avirori naaka apaniroini
2S=work-EP-REGR-EP-ICPL-IRR=EMPH=CT.F you
I
solo
aNtavai-t-atsi=ni
naaka am-atsi=ni
iroka naaka kamitaNt-atsi=ni
work-EP-STAT=REL I
bring-STAT=REL DEM I
buy-STAT=REL
iroka naakataki
DEM FOC
‘You also work then please, I am the only one who will work, I am the one who
will bring [food], I am the only one who will buy this, do that.’
42.
Ironyaaka nokantahetziri notomipaye, ‘Airo pooyakotana naaka, naaka
nonkamakerika, ninka pooyakotanaririka?’
ironyaaka no=kaNt-a-he-tz-i=ri
no=tomi-paye
airo
now
1SG.A=say-EP-PL-EP-REAL=3m.O 1SG.poss=son-PL NEG.IRR
p=ooy-ako-t-an-a
naaka naaka no=N-kam-ak-e=rika
2S=expect-APPL-EP-DIR-REAL I
I
1SG.S=IRR-die-PRF-IRR=COND
niNka p=ooy-ako-an-a=ri=rika
who 2A=expect-APPL-DIR-REAL=REL=COND
‘I said to my sons, ‘Don’t expect from me [help], when I die, from whom will you
expect [help]?’
43.
‘Arima pooyakotanahina, naaka, airo.’
ari=ma p=ooy-ako-t-an-ah-e=na
naaka airo
PP=DUB 2A=expect-APPL-EP-DIR-REGR-IRR=1SG.O I
NEG.IRR
‘In case, you expect [something] from me, I won’t [help you].’
44.
Ironyaaka nantavaitziniri, notyankiniri aisatzi notomi saikatsiri Irimashiki.
ironyaaka n=aNtavai-tz-i=ni=ri
no=tyaNk-i=ni=ri
now
1SG.A=work-EP-REAL=3O=3m.O 1SG.A=send-REAL=3O=3m.O
aisatzi no=tomi
saik-atsi=ri
Irimashi-ki
also
1SG.poss=son be.at-STAT=REL Lima-LOC
‘I earned it [money] for him, I sent it to him, to my son who lives in Lima.’
309
45.
Okimita inkaranki ikaimakina ashiroki, ikantzi, ‘Aapa, tekatsitaki naaka
novashitsiteni, pintyankahaitena kapichiini naari.’
okimita
iNkaraNki i=kaim-ak-i=na
ashiro-ki i=kaNt-tz-i
for.example recently 3m.A=call-PRF-REAL=1SG.O steel-LOC 3m.S=say-EP-REAL
aapa tekatsi-t-ak-i
naaka no=vashitsi-te-ni
Dad NEG.EXIST-EP-PRF-REAL I
1SG.poss=money-poss-DIM
pi=N-tyaNk-ah-a-it-e=na
kapichiini naari
2A=IRR-send-REGR-EP-ICPL-IRR=1SG.O little
I
‘For example, recently he called me on the phone, and said, ‘Dad, I don’t have
money, send me some money.’
46.
Osheteniranki nohatapaintzi kirinkanta, ipinaitakina kapichiini, iritaki
notyankakiniri inkaranki.
osheteniraNki no=ha-t-apaiNt-tz-i
kiriNka=Nta
yesterday
1SG.S=go-EP-once-EP-REAL downriver=DEM
i=pina-ai-t-ak-i=na
kapichiini
3m.A=pay-IMP.P-EP-PRF-REAL=1SG.O some
iritaki
FOC
no=tyaNk-ak-i=ni=ri
1SG.A=send-PRF-REAL=3O=3m.O
‘Yesterday I went downriver, they paid me some money, this is what I sent to
him.’
47.
Irotaki nokantahetziri nosaikahetzika, ‘Aparopaye aminahetzi [antavairintsi].’
irotaki no=kaNt-a-he-tz-i=ri
no=saik-a-he-tz-i=ka
FOC 1SG.A=say-EP-PL-EP-REAL=3m.O 1SG.S=be.at-EP-PL-REAL=ADV
aparo-paye amin-a-he-tz-i
aNtavai-rintsi
one-PL
look.for-EP-PL-EP-REAL work-NMZ
‘This is what I say to all of them living [in my house], ‘Each [person] is looking
for work.’
48.
Notomi pashini iyotakiro irirori, yaanaki iriikite irirori.
no=tomi
pashini i=yo-t-ak-i=ro
irirori
1SG.poss=son other 3m.A=know-EP-PRF-REAL=3n.m.O he
y=a-an-ak-i
i=iriki-te
irirori
3m.S=take-DIR-PRF-REAL 3m.poss=money-poss he
‘My other son knows it, he earns his own money.’
310
49.
Irotaki okaratzi nokamantzimiri maaroni, okaratzi nantziri kitaiteripayeka.
irotaki o=kara-tz-i
no=kamaNt-tz-i=mi=ri
FOC
3n.m.S=COP.C-EP-REAL 1SG.A=inform-EP-REAL=2O=REL
maaroni o=kara-tz-i
n=aNt-tz-i=ri
kitaiteri-paye=ka
all
3n.m.S=COP.C-EP-REAL 1SG.S=do-EP-REAL=REL day-PL=DEM
‘This is all that I was telling you, all that I do these days.’
50.
Nonivitari naari, nonivipiratari ikantziro ‘intsikiyatyero piroota’.
no=nivi-t-a=ri
naari no=nivi-pira-t-a=ri
1SG.A=like-EP-REAL=REL I
1SG.A=like-AUG-EP-REAL=REL
i=kaNt-tz-i=ro
i=N-tsikiy-aty-e=ro
piroota
3m.A=say-EP-REAL=3n.m.O 3m.A=IRR-kick-PROG-IRR=3n.m.O ball
‘What I like, what I like most, they call it ‘kicking the ball’ [football].’
51.
Maatsi irora ikantzi ‘ampinatavakaya’ kiriiki.
tzimatsi irora i=kaNt-tz-i
a=N-pina-t-av-ak-a-ia
EXIST
kiriiki
DEM 3m.S=say-EP-REAL 1PL.S=IRR-pay-EP-RCP-PRF-EP-IRR.REFL money
‘There is this, what they call ‘a money bet’.
52.
Kinkivarivitakana, shiyachana irohatzi, kametsa notsikiyiro nyaatsamentotsi,
inintziro evankarikapaye, ipinkatsatai.
kiNkivari-t-ak-a=na
shiy-acha=na
irohatzi kametsa
be.old-EP-PRF-REAL=1SG.O run-STAT=1SG.O until
well
no=tsikiy-i=ro
nyaatsa-meNto-tsi i=niNt-tz-i=ro
1SG.A=kick-REAL=3n.m.O play-NMZ-ABS
3m.A=like-EP-REAL=3n.m.O
evaNkari-paye i=piNkatsa-t=ai
youth-PL
3m.A=obey-EP=1PL.O
‘I am old [but] still [can] run, I kick the ball well, young people like it, they
respect us [me].’
53.
Arika inyavakina nonyaatsatatyeya kantanaki, ‘Te, iyora iyotziro inyaatsata.’
arika i=ny-av-ak-e=na
no=nyaatsa-t-aty-eya
when 3m.A=see-DIR-PRF-IRR=1SG.O 1SG.S=play-EP-PROG-IRR.REFL
kaNt-an-ak-e
te
iyora i=yo-tz-i=ro
i=nyaatsa-t-a
say-DIR-PRF-IRR NEG DEM 3m.A=know-EP-REAL=3n.m.O 3m.S=play-EP-REAL
‘When they see me play [with them], they say, ‘No, this [guy] knows how to
play.’
311
54.
‘Naakamaha te noyopirotero.’
naaka=ma=ha
te
no=yo-piro-t-e=ro
I=DUB=EMPH NEG.REAL 1SG.A=know-AUG-EP-IRR=3n.m.O
‘In contrast, we don’t know [how to play].’
55.
Nokantziri, ‘Tema aviroka evankari pinatzi, naaka te nevankarite.’
no=kaNt-tz-i=ri
tema
aviroka evaNkari pi=na-tz-i
naaka
1SG.A=say-EP-REAL=3m.O although you
youth
2S=be-EP-REAL I
te
n=evaNkari-t-e
NEG.REAL 1SG.S=be.young-EP-IRR
‘I tell them, ‘Although you are young, I am not young.’
56.
Aisatzi oshekitaki anyaatsari, kapicheeni kiriiki impahai.
aisatzi osheki-t-ak-i
a=nyaatsa-ri
kapicheeni kiriiki
also
many-EP-PRF-REAL 1PL.poss=play-NMZ little
money
i=N-p-ah=ai
3m.A=give-REGR=1PL.O
‘Also, there are many games, [but] they pay us little money.’
57.
Ari ankamitantakyari asuka, tanta, avamanyari okitaitamani.
ari a=N-kamitaNt-ak-ia=ri
asuka taNta a=v-aman-ia=ri
PP 1PL.A=IRR-buy-PRF-IRR=3mO sugar bread 1PL.A=eat-early-IRR=3m.O
o=kitait-aman-ni
3n.m.S=be.morning-early-ADV
‘[With this] we will buy sugar and bread, we will eat them early in the morning.’
58.
Irotaki okaratzi nantziri.
irotaki o=kara-tz-i
n=aNt-tz-i=ri
FOC 3n.m.S=COP.C-EP-REAL 1SG.A=do-EP-REAL=REL
‘This is all that I do.’
59.
Ironyaaka kitaiteripayeka aisatzita pashinipaye antavairintsi, nantzi
namitakovaitantzi, namiri kireeki.
ironyaaka kitaiteri-paye=ka aisatzi=ta pashini-paye aNtavai-rintsi
now
day-PL=DEM also=OPT other-PL
work-NMZ
n=aNt-tz-i
n=amitako-vai-t-aNt-tz-i
1SG.S=do-EP-REAL 1SG.S=help-DUR-EP-APPL.REAS-EP-REAL
n=am-i=ri
kireeki
1SG.A=bring-REAL=3n.m.O money
‘These days also [I do] other jobs helping [friends], and bring money.’
312
60.
Te nonimopirotya noshinkipirotya.
te
no=nimo-piro-t-ia
no=shiNki-piro-t-ia
NEG.REAL 1SG.S=like-AUG-EP-IRR 1SG.S=get.drunk-AUG-EP-IRR
‘I don’t like to get drunk.’
61.
Arika impataikena, arika ontsitonkapahena kapicheeni, hatahena, onkantanaha
nonintziro naari nantavairi.
arika i=N-pataik-e=na
arika o=N-tsitoNK-ap-ah-e=na
when 3m.A=IRR-treat-IRR=1SG.O when 3n.m.A=IRR-hit-DIR-REGR-IRR=1S.O
kapicheeni ha-t-ah-e=na
o=N-kaNt-an-ah-a
little
go-EP-REGR-IRR=1SG.O 3n.m.S=IRR-happen-DIR-REGR-REAL
no=niNt-tz-i=ro
naari n=antavai-ri
1SG.A=want-EP-REAL=3n.m.O I
1SG.poss=work-NMZ
‘When treat me, when it [the alcohol] hits me, I leave because I like my job.’
62.
Ari okaratzi ironyaaka nokinkitsatavairi, paasonki.
ari o=kara-tz-i
ironyaaka no=kiNkitsa-t-av-a-i=ri
PP 3n.m.S=COP.C-EP-REAL now
1SG.A=tell-EP-DIR-EP-REAL=REL
paasoNki
thank.you
‘This is all now that I tell them, thank you.’
B.2
Song Kovacheri ‘Gray-necked wood-rail (Aramides cajanea)’ by Ernesto Manchi
Lopez (Pampa Michi)
1.
Nonyanananyanatakeri noyovancheritera.
no=nyanananyan~a-t-ak-i=ri
no=yovancheri=ra
1SG.A=see.DIR~REP-EP-PRF-REAL=3m.O 1SG.poss=wood-rail=DEM
‘I’ve seen my word-rail.’
2.
Ishiyashiyatanaka isavoroshiteki.
i=shiyashiya~a-t-an-ak-a
i=savoro-shi-te-ki
3m.S=run~REP-EP-DIR-PRF-REAL 3m.poss=reed-CL:small.thin.flat-poss-LOC
‘He ran to the grassy area of [wild cane].
3.
Nonyananonyanatakeri noyovancheritera.
no=nyananonyan~a-t-ak-i=ri
no=yovancheri=ra
1SG.A=see.DIR~REP-EP-PRF-REAL=3m.O 1SG.poss=wood-rail=DEM
‘I’ve seen my wood-rail.’
313
4.
Ishiyashiyatanakara isavoroshiteki.
i=shiyashiya~a-t-an-ak-a=ra
i=savoro-shi-te-ki
3m.S=run~REP-EP-DIR-PRF-REAL=ADV 3m.poss=reed-CL:small.thin.flat-poss-LOC
‘He ran to the grassy area of wild reed.
5.
Kametsa ikanta, yaminaminatzira ishaninkaperori, ishaninkaperori.
kametsa i=kaNt-a
y=aminamin~a-tz-i=ra
handsome 3m.S=be-REAL 3m.S=look.for~REP-EP-REAL=ADV
i=shaniNka-pero-ri
i=shaniNka-pero-ri
3m.poss=fellowman-AUG-NMZ 3m.poss=fellowman-AUG-NMZ
‘He is handsome, he is looking for his true mate, his true mate.’
6.
Aritakitya noyovancheritera yaminayaminatzi ishaninkaperori.
aritaki=tya no=yovancheri=ra
y=aminayamin~atz-i
PP=EMPH 1SG.poss=wood-rail=DEM 3m.S=look.for~PROG-REAL
i=shaniNka-pero-ri
3m.poss=fellowman-AUG-NMZ
‘That my wood-rail is looking for his true mate is the case.’
7.
Aritaitakima, inyanyatakiri ishaninkaperori, irirori ikanta.
aritarita~ak-i=ma
i=nyanya~t-ak-e=ri
PP~PRF-REAL=DUB 3m.A=look.for~EP-PRF-IRR=3m.O
i=shaniNka-pero-ri
irirori i=kaNt-a
3m.poss=fellowman-AUG-NMZ he
3m.S=be-REAL
‘That he will look for his true mate will be the case, he is.’
8.
Aritaki ishaninka yaminaminatzira ishaninkaperori noyovancheritera.
aritaki i=shaniNka
y=aminamin~atz-i=ra
PP
3m.poss=fellowman 3m.S=look.for~PROG-REAL=ADV
i=shaniNka-pero-ri
no=yovancheri-te=ra
3m.poss=fellowman-AUG-NMZ 1SG.poss=wood-rail-poss=DEM
‘That he is looking for his true mate is the case, my wood-rail.’
314
B.3
Riddle by Delia Rosas Rodríguez (Bajo Marankiari)
Antakiroki kenashiri, nonatzi intsompoiki nokitamarotatzi, tsika nopaitari?
aNtakiro-ki kenashiri no=na-tz-i
intsoNpoi-ki
outside-LOC green
1SG.S=be-EP-REAL inside-LOC
no=kitamaro-t-atz-i
tsika no=pai-t-a=ri
1SG.S=be.white-EP-PROG-REAL WH 1SG.S=be.called-EP-REAL=REL
‘I am green outside; I am white inside, what am I called?’ (Intsipa. Guava.)
B 4.
Conversation by Victoria Manchi de Martin and Raul Martin Bernata (Pampa
Michi)
1.
Raul: Shiteniranki ahataki Tsiirsihiki, naanakimi, pikimoshiritaki.
shiteni=raNki a=ha-t-ak-i
Tsiirisi-ki
n=a-an-ak-i=mi
day=ADV.T 1PL=go-EP-PRF-REAL La Merced-LOC 1SG.A=take-DIR-PRF-REAL=2O
pi=kimoshiri-t-ak-i
2S=be.glad-EP-PRF-REAL
‘Yesterday we went to La Merced, I took you along, you’re happy [now].’
2.
Victoria: Ehe, nokimoshiritaki, paanakinaranki.
ehe no=kimoshiri-t-ak-i
p=a-an-ak-i=na=raNki
yes 1SG.S=be.glad-EP-PRF-REAL 2A=take-DIR-PRF-REAL=1SG.O=ADV.P
‘Yes, I was glad when you took me along the other day.’
3.
Raul: Te okameestate ovari, ahatzi naari katsitanaki nomotya, aviroka, te?
te
o=kameetsa-t-e
ovari ahatzi naari katsit-an-ak-i
NEG.REAL 3n.m.S=be.good-EP-IRR food
also
I
feel.pain-DIR-PRF-REAL
no=motya
aviroka te
1SG.poss=stomach you
NEG
‘The food wasn’t good, also I had pain in my stomach, and you, no [you don’t
have the pain]?’
4.
Victoria: Naaka te nokatsite, ironyaaka airo ahatahi.
naaka te
no=katsit-e
ironyaaka airo
I
NEG.REAL 1SG.S=feel.pain-IRR now
‘I didn’t feel anything, and we are not going.’
a=ha-t-ah-i
NEG.IRR 1PL.S=go-EP-REGR-REAL
315
5.
Raul: Tsame ironyaaka ahate avapaintya, chookikinta Mishaja, te pininte?
tsame
ironyaaka a=ha-t-e
a=v-apaiNt-ia
chooki-ki=Nta
come.on now
1PL=go-EP-IRR 1PL=eat-once-IRR sister-LOC=DEM
Mishaja
te
pi=niNt-e
restaurant’s.name NEG.REAL 2S=want-IRR
‘Let’s go eat at the sister’s restaurant ‘Mishaja’, don’t you want to go?’
6.
Victoria: Ninka ovakahainika?
niNka ov-ak-ah=ai=ni=ka
who
eat-CAUS-REGR=1PL.O=REL=Q
‘Who’s going to invite [feed] us?’
7.
Raul: Naaka, naaka ovakaimini.
naaka naaka ov-ak-a-e=mi=ni
I
I
eat-CAUS-REGR-IRR=2O=REL
‘I am the one who’s inviting you.’
8.
Victoria: Tzimatsi piirikite?
tzimatsi pi=iriki-te
EXIST 2poss=money-poss
‘Do you have money [to go eat out]?’
9.
Raul: Aririka nopimantakiro.
aririka no=pimaNt-ak-e=ro
.
when 1SG.A=sell-PRF-IRR=3n.m.O
‘When I sell it [the fruit then I’ll have it].’
10.
Victoria: Aririka apimantakiro chochoki, airo opimi, tekatsite oirikite.
aririka a=pimaNt-ak-e=ro
chochoki airo
when 1PL.A=sell-PRF-IRR=3n.m. fruit
NEG.IRR
o=p-i=mi
tekatsi-t-e
o=iriki-te
3n.m.A=give-REAL=2O NEG.EXIST-EP-IRR 3n.m.poss=money-poss
‘When we sell the fruit, she’s not going to pay you, she [the wholesaler] won’t
have the money.’
11.
Raul: Ari noyatanakiro.
ari n=oya-t-an-ak-e=ro
PP 1SG.A-follow-EP-DIR-PRF-IRR-3n.m.O
‘I’ll follow her [I’ll nag her to give me the money].’
316
12.
Victoria: Airo nayimi chochoki, paashityaniro aririka ampitsaitsokitanakya,
ninka kamitantahi iroka?
chochoki p=ashi-t-ia=ni=ro
airo
n=ay-e=mi
NEG.IRR 1SG.A=take-IRR=2O fruit
2S=have-EP-IRR=3O=3n.m.O
aririka aNpitsai-ki-t-an-ak-ia
niNka kamitaNt-ah-e iroka?
when wrinkled-CL:small.round-DIR-PRF-IRR who buy-REGR-IRR DEM
‘I won’t pick fruit for you; [if] you harvest the fruit for her when the fruit is
wrinkled, who’s going to buy it?’
13.
Raul: Iro paitaranki, osamanivitakya ari opinatyai, ari naaki ironyaaka.
iro paitaraNki
osamani-vi-t-ak-ia
ari
but whachamacallit some.time.later-FRUS-EP-PRF-IRR PP
o=pina-t-aty=ai
ari
n=a-ak-e
ironyaaka
3n.m.S=pay-EP-PROG=1PL.O PP 1SG.S=harvest-PRF-IRR now
‘But whachamacallit, later on if she is going to pay us, I will harvest now.’
14.
Victoria: Osaikera, airo ahatzi ironyaaka, tema pitsonkanari piirikite.
osaikira airo
a=ha-tz-i
ironyaaka tema
tomorrow NEG.IRR 1PL.S=go-EP-REAL now
since
pi=tsoNk-an-ak-e=ri
pi=iriki-te
2poss=money-poss
‘Tomorrow. We are not going now because you ran out of money.’
2A=finish-DIR-PRF-REAL=REL
15.
Raul: Nopanakiro noharanikite.
no=p-an-ak-i=ro
no=hananiki-te
1SG.A=give-DIR-PRF-REAL=3n.m.O 1SG.poss=little.child-poss
‘I gave the money to my daughter.’
16.
Victoria: Irotakitya arimantetakya, okaratzi opimantari irotakima; aminirota
osaikera.
irotaki=tya
arima-aNt-e=ta=kya
o=kara-tz-i
FOC=EMPH maybe-APPL.REAS-IRR=OPT=EMPH 3n.m.S=CAP.C-EP-REAL
o=pimaNt-a=ri
irotaki=ma amin-e=ro=ta
osaikira
3n.m.S=sell-REAL=REL FOC=DUB see-IRR=3n.m.O=OPT tomorrow
‘That’s why it will probably be enough what you’ve sold, we’ll see tomorrow.’
17.
Raul: Paitaka anyantarori?
paita=ka a=ny-aNt-ia=ro=ri
WH=Q 1PL.A=see-APPL.REAS-IRR=3n.m.O=REL
‘What for [why] are we going to see [to wait]?’
317
18.
Victoria: Aririka opakina saikanatsiri, osaikira ari pihataki Puchariniki.
aririka o=p-ak-i=na
saik-an-atsi=ri
osaikira ari
when
3n.m.A=give-PRF-IRR=1SG.O be.at-DIR-STAT=REL tomorrow PP
pi=ha-t-ak-e
Pucharini-ki
2S=go-EP-PRF-IRR village’s.name-LOC
‘When she [the wholesaler] gives me the rest [of the money]; tomorrow you’ll go
to Pucharini.’
20.
Raul: Airo nohatzi tema nokosechatzi.
airo
no=ha-tz-i
tema
no=kosecha-t-atz-i
NEG.IRR 1SG.S=go-EP-REAL because 1SG.S=harvest-EP-PROG-REAL
‘I won’t travel [to Pucharini] because I am harvesting [fruit].’
21.
Victoria: He, he, airo pikosechatatzi.
he he airo
pi=kosecha-t-atz-i
yes yes NEG.IRR 2S=harvest-EP-PROG-REAL
‘Yes, yes (laughs), you won’t harvest.’
22.
Raul: Tekatsi notsipatyari.
tekatsi
no=tsipa-t-ia=ri
NEG.EXIST 1SG.S=accompany-EP-IRR=REL
‘I have no companion [to harvest fruit].’
23.
Victoria: Airo pishikitziro.
airo
pi=shiki-tz-i=ro
NEG.EXIST 2S=many-EP-REAL=3n.m.O
‘Don’t harvest a lot of fruit.’
24.
Victoria: Onintzi apite.
o=niNt-tz-i
apite
3n.m.S=want-EP-REAL two
‘She [the wholesaler] wants [to pay us] two [soles per crate of fruit].’
25.
Raul: Te okovi opinatero veinte soles.
te
o=kov-e
o=pina-t-e=ro
veinte soles
NEG.REAL 3n.m.S=want-IRR 3m.n.A=pay-EP-IRR=3n.m.O twenty soles
‘She doesn’t want to pay for it twenty soles.’
318
26.
Victoria: Tema ikantakiro Papi ipinatero veinte soles.
tema i=kaNt-ak-i=ro
Papi
i=pina-t-e=ro
may.be 3m.A=say-PRF-REAL=3n.m.O son’s.name 3m.A=pay-EP-IRR=3n.m.O
veinte soles
twenty Peruvian.currency
‘But Papi [son’s name] told her that he would pay her twenty soles [per crate of
fruit].’
27.
Raul: Te antaroite naye, te anyiri Papi akantavahiri.
te
aNtaro-ite n=ay-e
te
a=ny-e=ri
NEG.REAL big-PL
1SG.S=take-IRR NEG.REAL 1PL.A=see-IRR=3m.O
Papi
a=kaNt-av-ah-e=ri
son’s name 1PL.A=say-DIR-REGR-IRR=3m.O
‘I won’t harvest a lot, we won’t see our son to talk [about it].’
28.
Raul: Tema nokantzi te okante apaniroini airo akantziro, airo nokantzini
atziripaye.
tema no=kaNt-tz-i
apaniroini airo
a=kaNt-tz-i=ro
maybe 1SG.S=say-REAL solo
NEG.IRR 1PL.A=be.able-REAL=3n.m.O
airo
no=kaNt-tz-i-ni
NEG.IRR 1SG.S=say-REAL-EP-REAL-AUG
atziri-paye
person-PL
‘As I am saying that we can’t do it on our own; I won’t tell people [what to do].’
29.
Raul: Irotaki kaari itzimanta iirikite ipinatahiri yatziritepaye.
irotaki kaari
i=tzimaNt-a
i=iriki-te
FOC
NEG.COP 3m.S=have-REAL 3m.poss=poney-poss
i=pina-t-ah-e=ri
y=atziri-te-paye
3m.A=pay-EP-REGR-IRR=3m.O 3m.poss=person-poss-PL
‘That’s why that he has money to pay his staff is not the case.’
30.
Victoria: Tekatsi yaavahiri, airo apinatziri.
tekatsi
y=a-av-ah-i=ri
NEG.EXIST
airo
3m.S=take-DIR-REGR-REAL=REL NEG.IRR
a=pina-tz-i=ri
1PL.A=pay-EP-REAL=3m.O
‘There is no money for him to earn, and we won’t pay him.’
31.
Raul: Airo akantziri.
airo
a=kaNt-tz-i=ri
NEG.IRR 3n.m.S=say-REAL=3m.O
‘We shouldn’t say that to him.’
319
32.
Victoria: Ahatzi naari chapinki nokantavitari, ‘Pipinatziritaima kapichiini.’
ahatzi naari chapiNki no=kaNt-a-vi-t-a=ri
also I
recently 1SG.A=say-EP-FRUS-EP=3m.O
pi=pina-tz-i=ri=taima
kapichiini
2S=pay-EP-REAL=3m.O=DUB little
‘Last time I told him, ‘Pay them [your workers] a little bit [of what you paid them
before].’
33.
Raul: Iriro yaanakiri, yaminiri kapichikitaite irohatzi oshetyantahe, imakoryahari
sheteniki.
iriro y=a-an-ak-e=ri
y=amin-e=ri
kapichikitaite
he 3m.S=take-IRR=3m.O 3m.A=see-IRR=3m.O early.morning
irohatzi o=she-tya-t-aNt-ah-e
until
3n.m.S=be.day-almost-EP-APPL.REAS-REGR-IRR
i=makory-ak-atz-i=ri
sheteni-ki
3m.A=rest-CAUS-PROG-REAL=3m.S afternoon-LOC
‘He hires them, he looks after them from morning till night, he makes them take a
break in the afternoon.’
36.
Victoria: Iriro ikotsivintziro itsipatari Pito ironyaaka.
iriro i=kotsi-viNt-tz-i=ri
i=tsipa-t-a=ri
he
Pito
‘He cooks for them, together with Pito now.’
38.
Raul: Naaka oshiki naye, ovira nayetziri; aritapaki.
naaka oshiki n=ay-e
o=vir-a
I
much 1SG.S=take-IRR 3n.m.S=come.near-REAL
n=a-yi-tz-i=ri
ari-t-ap-ak-i
1SG.A=take-DIST-REAL=REL PP-EP-DIR-PRF-REAL
‘I’ll harvest a lot, I harvest this much; no more.’
39.
ironyaaka
3m.A=cook-BEN-EP-REAL=3m.O 3m.A=join-EP-REAL=3m.O person’s.name now
Victoria: Iroo pikante osankinatemita.
iroo pi=kaNt-e
o=saNkina-t-e=mi=ta
she 2S=say-IRR 3m.A=write-EP-IRR=2O=OPT
‘She will write down what you say to her.’
Este arquivo, fornecido pela autora em dezembro de 2010,
está disponível para download no seguinte endereço:
http://www.etnolinguistica.org/tese:mihas_2010
320
CURRICULUM VITA
Elena I. Mihas
Place of birth: Smolensk, Russia
Education
M.A. in English, University of Wisconsin-Milwaukee (UWM), May 2006
Major: Linguistics
B.A., Belgorod State Pedagogical Institute, Russia, magna cum laude, June 1984
Double major: History and Social Science, and English as a Foreign Language
Dissertation Title: Essentials of Ashéninka Perené Grammar
Awards, Fellowships, Grants
2009. Hans Rousing Foundation Endangered Languages Project, School of Oriental and
African Studies, London, UK, SG0002 “Description and Documentation of Ashéninka
Perené (Arawak)”.
2009. National Science Foundation DDIG #0901196 “Ashéninka Perené: Documentation
and grammar”.
2008. James A. Sappenfield Graduate Fellowship, University of Wisconsin-Milwaukee.
Graduate fieldwork
May-September 2010. Fieldwork in the villages of Pampa Michi, Villa Perené, and Bajo
Marankiari. Districts of Chanchamayo and Perené, Junin Province, Peru.
June-September 2009. Fieldwork in the villages of Pampa Michi, Villa Perené, and Bajo
Marankiari. Districts of Chanchamayo and Perené, Junin Province, Peru.
June 2008. Survey of potential research base communities of Ashéninka Perené speakers
in Districts of Chanchamayo and Perené, Junin Province, Peru.
Publications
Forthcoming, 2010. “Gender system in Ashéninka Perené.” UniverSOS, Revista de
Lenguas Indígenas y Universos Culturales 7.
Forthcoming, 2011. “Bodily-based conceptual metaphors in Ashéninka Perené myths and
folk stories”, in Endangered metaphors, Anna Idström, & Elisabeth Piirainen (eds.), John
Benjamins.
2010. “Orthography development in Ashéninka Perené”, Cadernos de Etnolinguistica
2(2). Available online at http://www.etnolinguistica.org/issue:vol2n2.
2010. “Nominal classification in Ashéninka", LSA 2010 Annual Meeting Extended
Abstracts. Available online at http : //www.elanguage.net/journal
s/index.php/lameeting/index.
2010. “High Definition Camera HDC-HC 100P/PC and HD Writer 2.6E High Definition
321
Image Management/Easy Editing Software” (co-authored with Jeffrey Loomis),
Language Documentation and Conservation 4:66-74.
2009. "Negation in Metta", Rice Working Papers in Linguistics 1:197-222.
2005. "Non-literal language in political discourse", Working Papers in Linguistics 5: 124139, Proceedings of the LSO Workshop in General Linguistics.
Teaching materials
2010. Ashéninka Perené collection. The Archive of the Indigenous languages of Latin
America. Available online at
http://www.ailla.utexas.org/search/collection.html?c_id=113.
Conference Presentations
“Nominal classification in Ashéninka (Arawak): Towards a multidimensional typology.”
Paper presented at the Linguistic Society of America Annual Meeting, Baltimore, January
7-10, 2010.
“Documentary Linguistics: A fieldtrip to Amazonia”, Talk given at the MAFFL
Colloquium, University of Wisconsin-Milwaukee, October 1, 2009.
“Semantic functions of the Kampan applicative -ako from a typological perspective.”
Paper presented at the Biannual Meeting of the Association for Linguistic Typology, July
23-26, 2009, University of California, Berkeley.
“Grammaticalization of the Kampan applicative –ako.”Paper presented at 12th
Workshop on American Indigenous Languages, University of California, Santa Barbara,
May 8-9, 2009.
“Areal dimensions in syncretisms”, with M. Noonan. Paper presented at the 7th
Biannual Meeting of the Association for Linguistic Typology, Paris, September 25-28,
2007.
“Non-literal language in political discourse.” Paper presented at the Third Workshop in
General Linguistics, University of Wisconsin-Madison, Madison, February 25, 2005.
Positions Held
2005-2010. Teaching Assistant, English Department, UWM. Courses taught: English
101, English 102, English 209
________________________________________________________________________
Major Professor